🏷

Classifications

ICD-O Morphology

Harrison's Part

◆

Subtypes

2

Conventional EHE (WWTR1-CAMTA1 fusion) NCIT:C177552

The conventional and by far the most common form, accounting for approximately 90% of EHE. Defined by the t(1;3)(p36.23;q25.1) WWTR1(TAZ)-CAMTA1 fusion. Morphology is cords, strands and nests of epithelioid endothelial cells in a myxohyaline stroma, typically without well-formed vascular channels. Nuclear CAMTA1 immunohistochemistry is the surrogate marker.

YAP1-TFE3-fused hemangioendothelioma NCIT:C177553

A minority molecular subset (about 5% of cases, reported from <5% up to ~10%) occurring in younger adults, with more overt vasoformative architecture, voluminous eosinophilic to vacuolated cytoplasm, strong nuclear TFE3 expression and negative CAMTA1. Reported survival is higher than in the WWTR1-CAMTA1 subset (86% vs 58-59% at 5 years) despite a high propensity for multifocal and metastatic presentation, but the difference did not reach statistical significance in the largest comparison (p = 0.095), so the prognostic separation should be treated as suggestive rather than established.

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"Although the disease specific survival rate at 5 years was 86% in TFE3-rearranged patients compared to 58% in CAMTA1-rearranged patients, the difference between groups did not reach statistical signficance (p = 0.095)."

Gives both the survival difference and the explicit statement that it was not statistically significant, which is why the subtype prognostic claim is hedged. Support is PARTIAL for this reason.

?

Discussions and Knowledge Gaps

4

Does EHE multifocality arise by true metastatic dissemination from a single clone, or by field-effect / synchronous multicentric transformation?

KNOWLEDGE GAP ehe_multifocality_clonality

Attached to

pathophysiology#Multifocal Dissemination and Serosal Involvement

Multifocality is present in about three quarters of cases at diagnosis and dominates management decisions. For multifocal *hepatic* disease the question is largely settled: separate lesions share an identical genomic breakpoint, i.e. clonal metastatic dissemination within one organ. What remains open is whether the same holds for the far commoner multi-compartment pattern (e.g. synchronous lung and liver, or lung and bone), where clonality has not been demonstrated at scale and where the supporting argument is still only circumstantial - outcome data showing no survival difference between multifocal and metastatic disease. The answer determines whether multi-organ multifocal EHE should be staged and treated as metastatic disease.

Proposed experiments

Cross-compartment fusion-breakpoint sequencing of synchronous lesions

ehe_multiregion_breakpoint_sequencing

Extend the published intrahepatic shared-breakpoint analysis to synchronous lesions in different compartments (lung plus liver, lung plus bone) using whole-genome or targeted breakpoint sequencing, to test whether multi-organ multifocality is also clonal rather than multicentric.

Phylogenetic reconstruction from secondary alterations

ehe_synchronous_lesion_phylogeny

Comparison of secondary alterations such as CDKN2A status across paired synchronous lesions to reconstruct clonal relationships.

Does the CTGF-integrin-RAS-MAPK arm mapped in TAZ-CAMTA1-transformed NIH3T3 fibroblasts actually operate in human EHE endothelium?

HUMAN MODEL MISMATCH ehe_ctgf_mapk_model_fidelity

Attached to

pathophysiology#CTGF-Integrin-RAS-MAPK Signaling

The entire CTGF/integrin/RAS-MAPK dependency - and with it the rationale for the SARC033 trametinib trial - was established in TAZ-CAMTA1-transformed NIH3T3 mouse fibroblasts and their xenografts. EHE is a tumour of endothelial lineage, and the two independent genetic mouse models of the disease both act in the endothelial compartment. Whether a fibroblast-based transformation assay faithfully reports the signalling wiring of human EHE endothelium is unresolved, and this is precisely the kind of mismatch that can make a mechanistically compelling target fail clinically. This is a HUMAN_MODEL_MISMATCH rather than a KNOWLEDGE_GAP because the evidence exists and is strong - it is its translational validity that is in question.

Proposed experiments

Re-test CTGF-RAS-MAPK dependence in endothelial and patient-derived EHE models

ehe_ctgf_mapk_in_endothelial_models

Repeat the CTGF knockdown, integrin-blockade and MEK-inhibition experiments in TAZ-CAMTA1-expressing endothelial cells, the established EHE cell lines and the patient-derived xenograft, rather than in NIH3T3 fibroblasts.

MAPK pathway activity in human EHE tissue

ehe_mapk_activity_in_human_tumours

Quantify phospho-ERK and CTGF expression in primary human EHE specimens, and correlate with trametinib response among SARC033 participants, to confirm the pathway is active in the human disease.

Why do patients with serosal effusion respond so poorly to sirolimus, and is there a distinct biology underlying effusion-positive EHE?

KNOWLEDGE GAP ehe_serosal_effusion_sirolimus_resistance

Attached to

pathophysiology#Multifocal Dissemination and Serosal Involvement

Serosal effusion is the most reproducible adverse prognostic factor in EHE and specifically predicts sirolimus failure, but no molecular correlate has been identified. Whether effusion marks a distinct tumour biology, a pharmacokinetic barrier, or simply advanced disease burden is unresolved, and the answer determines whether a different therapeutic class is needed for this subgroup.

Proposed experiments

Comparative profiling of effusion-positive versus effusion-negative EHE

ehe_effusion_comparative_profiling

Transcriptomic and secondary-mutation profiling of matched effusion-positive and effusion-negative tumours to identify a molecular correlate of the adverse phenotype.

Pharmacodynamic assessment of mTORC1 inhibition in effusion-positive disease

ehe_effusion_mtorc1_pharmacodynamics

Measurement of intratumoural sirolimus exposure and phospho-4E-BP1 as a pharmacodynamic readout, to distinguish drug-delivery failure from target-independent resistance.

Should YAP1-TFE3-fused hemangioendothelioma be classified as a separate disease entity rather than an EHE subtype?

KNOWLEDGE GAP ehe_yap1_tfe3_nosology

Attached to

pathophysiology#YAP1-TFE3 Fusion Oncogene Formation

The YAP1-TFE3 subset differs from conventional EHE in driver fusion, architecture, immunophenotype, patient age and survival, and expert groups have proposed splitting it out. MONDO:0015523 currently subsumes both. The nosology decision affects trial eligibility and prognostic communication, and turns on whether the two converge on a common downstream transcriptional program.

Proposed experiments

Comparative TEAD versus MiT target chromatin profiling

ehe_tead_vs_mit_chromatin_profiling

Direct comparative transcriptomic and chromatin-occupancy profiling of WWTR1-CAMTA1 versus YAP1-TFE3 tumours to determine whether the two fusions converge on a shared downstream program.

Fusion-stratified prospective outcome analysis

ehe_fusion_stratified_outcome_analysis

Pooled prospective outcome analysis stratified by fusion; the EURACAN registry (NCT06408441) is positioned to deliver this.

⚙

Pathophysiology

9

WWTR1-CAMTA1 Fusion Oncogene Formation

A somatic t(1;3)(p36.23;q25.1) translocation fuses the 5' portion of WWTR1 (encoding the Hippo effector TAZ) on 3q25 to the 3' portion of CAMTA1 on 1p36. The chimeric gene is driven by the WWTR1 promoter, which is highly active in endothelial cells, so CAMTA1 sequence - normally expressed only in brain - is inappropriately activated by a promoter-switch mechanism. The fusion retains the TAZ 14-3-3-binding and WW domains together with the CAMTA1 TIG DNA-binding domain, ankyrin repeats and IQ calmodulin-binding motifs.

endothelial cell CL:0000115

WWTR1 hgnc:24042 CAMTA1 hgnc:18806

Downstream

Constitutive Nuclear TAZ-TEAD Transcriptional Activation The fusion protein escapes Hippo-pathway cytoplasmic restraint and acts as a constitutively active TAZ.

Show evidence (2 references)

PMID:21885404 SUPPORT Human Clinical

"This WWTR1/CAMTA1 gene fusion is under the transcriptional control of the WWTR1 promoter and encodes a putative chimeric transcription factor that joins the amino terminus of WWTR1, a protein that is highly expressed in endothelial cells, in-frame to the carboxyl terminus of CAMTA1, a protein..."

The discovery paper establishes the fusion architecture and the promoter-switch mechanism that drives inappropriate CAMTA1 expression in endothelium.

PMID:21584898 SUPPORT Human Clinical

"Our results demonstrate the presence of a WWTR1-CAMTA1 fusion in all EHE tested from bone, soft tissue, and visceral location (liver, lung) in keeping with a unique and specific pathological entity."

Independent confirmation that the fusion is consistently present across anatomic sites and grades, establishing it as the disease-defining lesion.

YAP1-TFE3 Fusion Oncogene Formation

In a minority of tumours a YAP1-TFE3 fusion replaces WWTR1-CAMTA1. YAP1 exon 1 is fused to TFE3 exon 4, retaining the TFE3 bHLH-leucine-zipper DNA-binding and transactivation domains while discarding the YAP1 S127 14-3-3 site, WW domain and C-terminal transactivation domain. YAP1 therefore supplies a strong endothelial promoter that misexpresses an otherwise intact MiT-family transcription factor. These tumours are WWTR1-CAMTA1-negative, CAMTA1-immunonegative and strongly TFE3-positive.

YAP1 hgnc:16262 TFE3 hgnc:11752

DNA-binding transcription factor activity GO:0003700 ⚠ ABNORMAL

Downstream

Epithelioid Endothelial Neoplastic Proliferation Misexpressed TFE3 drives an alternative oncogenic transcriptional program producing a morphologically distinct epithelioid vascular tumour.

Show evidence (2 references)

PMID:23737213 SUPPORT Human Clinical

"In summary, we are reporting a novel subset of EHE occurring in young adults, showing a distinct phenotype and YAP1-TFE3 fusions."

Defines the alternative fusion driver and its distinct clinicopathologic subset.

PMID:34381186 SUPPORT Human Clinical

"All cases were confirmed by molecular testing to harbor YAP1-TFE3 gene fusions: majority with YAP1 exon 1 fused to TFE3 exon 4 (88%), or less commonly, TFE3 exon 6 (12%)."

Establishes the recurrent breakpoint structure of the YAP1-TFE3 fusion in a genetically confirmed multi-institutional series.

Constitutive Nuclear TAZ-TEAD Transcriptional Activation

Normally, Hippo-pathway LATS kinases phosphorylate TAZ, promoting 14-3-3 binding and cytoplasmic sequestration. TAZ-CAMTA1 loses the C-terminal regulatory elements and gains CAMTA1-mediated nuclear localisation, so it is largely relieved of that restraint and behaves as a constitutively activated TAZ. Importantly, the restraint is attenuated rather than abolished: LATS1/2 deletion and mutation of the LATS phosphosites still increase the fusion's nuclear localisation and target-gene output, so TAZ-CAMTA1 remains susceptible to Hippo regulation - which is why interventions that drive YAP/TAZ phosphorylation and cytoplasmic sequestration (e.g. statins) can inhibit it. The nuclear fusion drives TEAD-dependent transcription; disruption of the fusion-TEAD interaction or expression of a dominant-negative TEAD blocks transformation, identifying this node as the central effector of the disease and the primary fusion-directed drug target.

hippo signaling GO:0035329 ↓ DECREASED positive regulation of transcription by RNA polymerase II GO:0045944 ↑ INCREASED

transcription coactivator activity GO:0003713 ↑ INCREASED

Downstream

Endothelial Cell Reprogramming Sustained TEAD-dependent transcription rewires the endothelial gene expression program.

CTGF-Integrin-RAS-MAPK Signaling CTGF is a tumorigenic transcriptional target of the fusion, relaying its output into RAS-MAPK signalling.

mTORC1 Pathway Activation YAP/TAZ activity converges on mTOR complex 1, the biologic rationale for mTOR inhibition in EHE.

Show evidence (3 references)

PMID:36598859 SUPPORT Human Clinical

"This fusion gene has been observed in almost all reported EHE cases and functions as a constitutively activated TAZ."

States directly that the fusion behaves as a constitutively active TAZ, the defining molecular consequence modelled by this node.

PMID:33766984 SUPPORT Model Organism

"disruption of the TAZ-CAMTA1-TEAD interaction or ectopic expression of a dominant negative TEAD in vivo inhibits TAZ-CAMTA1-mediated transformation"

Functional in vivo demonstration that TEAD engagement is required for fusion-driven transformation, establishing TAZ-TEAD activity as the central effector.

PMID:33766984 SUPPORT In Vitro

"We conclude that, despite its gain-of-function activity, TAZ-CAMTA1 is still susceptible to regulation by Hippo signaling."

Qualifies the "constitutive" framing: the fusion escapes most but not all Hippo restraint, which is the basis for pharmacological strategies that restore cytoplasmic sequestration. Support is PARTIAL because it tempers rather than simply confirming the node's central claim.

mTORC1 Pathway Activation

YAP/TAZ activity - and hence the EHE fusion oncoproteins - converges on activation of mTOR complex 1. This is the biologic rationale for sirolimus, the systemic agent with the strongest disease-specific activity signal in EHE. Sirolimus acts on this node, inhibiting phosphorylation of the mTORC1 substrate 4E-BP1 and, downstream of that, suppressing the ATF4/ATF5-driven release of GDF-15 that tracks disease aggressiveness.

TORC1 signaling GO:0038202 ↑ INCREASED

Downstream

Epithelioid Endothelial Neoplastic Proliferation mTORC1-driven cap-dependent translation supports tumour cell growth and proliferation.

Show evidence (2 references)

PMID:33107994 SUPPORT Other

"In vitro studies have demonstrated that YAP/TAZ, the oncogenic fusions of which are the molecular hallmark of EHE, are involved in activation of mTOR complex 1."

States the mechanistic link from the EHE fusion effectors to mTORC1. Evidence source is OTHER because this is an editorial commentary summarising the in vitro literature.

PMID:39283723 SUPPORT In Vitro

"This reduction was attributed to the drug-induced inhibition of phosphorylation/activation of 4E-BP1 and subsequent downregulation of the GDF-15 transcription factors ATF4 and ATF5."

Demonstrates that sirolimus engages this node in EHE models, with 4E-BP1 phosphorylation as the proximal mTORC1 readout.

Endothelial Cell Reprogramming

Expression of TAZ-CAMTA1 in endothelial cells initiates an angiogenic and regenerative-like transcriptional program and is by itself sufficient to generate vascular tumours with the distinctive features of EHE. Two independent conditional mouse models - endogenous-locus knock-in and endothelial-specific expression - reproduce human EHE, and inhibiting the fusion causes established tumours to regress, establishing continued oncogene dependence (oncogene addiction).

blood vessel endothelial cell CL:0000071

angiogenesis GO:0001525 ↑ INCREASED

Downstream

Epithelioid Endothelial Neoplastic Proliferation Reprogrammed endothelial cells form the epithelioid neoplastic population of EHE.

Show evidence (3 references)

PMID:33766984 SUPPORT Model Organism

"we show that TAZ-CAMTA1 expression in endothelial cells is sufficient to drive the formation of vascular tumors with the distinctive features of EHE, and inhibition of TAZ-CAMTA1 results in the regression of these vascular tumors"

Demonstrates both sufficiency in the endothelial compartment and ongoing dependence of established tumours on the fusion.

PMID:33766984 SUPPORT Model Organism

"We show that TAZ-CAMTA1 initiates an angiogenic and regenerative-like transcriptional program in endothelial cells"

Characterises the reprogrammed transcriptional state of the endothelial cell of origin.

PMID:33766982 SUPPORT Model Organism

"These mice develop EHE tumors that are indistinguishable from human EHE clinically, histologically, immunohistochemically, and genetically."

An independent endogenous-locus mouse model confirms that the fusion alone recapitulates human EHE with exquisite tissue specificity.

CTGF-Integrin-RAS-MAPK Signaling

Connective tissue growth factor (CTGF/CCN2) is a tumorigenic transcriptional target of TAZ-CAMTA1. Secreted CTGF binds integrin alphaIIbbeta3, sustaining anchorage-independent proliferation, and drives RAS and MAPK signalling; CTGF knockdown reduces pathway activity and MEK inhibition with PD 0325901 or trametinib abolishes fusion-driven anchorage-independent growth and xenograft expansion. This is the disease-specific route by which a Hippo-pathway lesion converges on the conserved constitutive-mitogenic-signalling hallmark node, and it supplied the preclinical rationale for the phase II trametinib trial. Model caveat: this arm was mapped in TAZ-CAMTA1-transformed NIH3T3 fibroblasts and their xenografts, not in native human EHE endothelium, so its fidelity to the endothelial disease is an open question (see the linked HUMAN_MODEL_MISMATCH discussion).

Ras protein signal transduction GO:0007265 ↑ INCREASED ERK1 and ERK2 cascade GO:0070371 ↑ INCREASED anoikis GO:0043276 ↓ DECREASED

Downstream

Epithelioid Endothelial Neoplastic Proliferation RAS-MAPK output sustains anchorage-independent tumour cell proliferation.

Show evidence (2 references)

PMID:35443056 SUPPORT In Vitro

"We identified connective tissue growth factor (CTGF) as a tumorigenic transcriptional target of TC."

Identifies CTGF as the direct tumorigenic transcriptional target of the fusion oncoprotein.

PMID:35443056 SUPPORT In Vitro

"Collectively, our results demonstrated that CTGF and the Ras-MAPK signaling cascade are essential for TC-mediated tumorigenesis."

Establishes RAS-MAPK dependence, the conformance basis for the sustaining-proliferative-signaling hallmark node.

Epithelioid Endothelial Neoplastic Proliferation

The reprogrammed endothelial cells form the characteristic lesion: cords, strands and nests of epithelioid cells embedded in a myxohyaline (myxochondroid to sclerotic) stroma, with primitive intracytoplasmic vascular lumina ("blister cells") rather than well-formed vascular channels. Tumours retain endothelial differentiation (CD31/PECAM1 and ERG positive) and grow in an infiltrative pattern in liver, lung, bone and soft tissue.

neoplastic epithelioid endothelial cell CL:0000115

cell population proliferation GO:0008283 ↑ INCREASED

liver UBERON:0002107 lung UBERON:0002048 bone tissue UBERON:0002481

Downstream

Multifocal Dissemination and Serosal Involvement The tumour is frequently multifocal at diagnosis and disseminates to lymph nodes, other viscera and serosal surfaces.

Show evidence (1 reference)

PMID:32913529 SUPPORT Human Clinical

"Histologically, EHE typically showed a mucous hyaline or cartilaginous stroma, often forming a primitive vascular lumen, and expressed vascular endothelial markers."

Describes the defining tissue-level architecture and retained endothelial phenotype of the neoplastic proliferation.

CDKN2A Loss and Tumour Progression

Roughly half of EHEs carry a secondary genetic alteration beyond the defining fusion; loss of the CDKN2A tumour suppressor is the most common. In a conditional mouse model, adding Cdkn2a knockout to the TAZ-CAMTA1 allele produced earlier tumour-related morbidity and mortality and enhanced tumour cell proliferation, identifying CDKN2A loss as a progression event rather than an initiating one. This is the disease-specific substitution for the conserved tumour-suppressor-inactivation trigger of the evading-growth-suppressors hallmark.

CDKN2A hgnc:1787

cell cycle checkpoint signaling GO:0000075 ↓ DECREASED negative regulation of cell cycle GO:0045786 ↓ DECREASED

Downstream

Epithelioid Endothelial Neoplastic Proliferation Loss of the CDKN2A-enforced cell-cycle checkpoint removes an antiproliferative brake, enhancing tumour cell proliferation. This is the disease-specific instance of the module edge Tumor Suppressor Inactivation to Loss of Cell-Cycle Checkpoint Control.

Show evidence (3 references)

PMID:36598859 SUPPORT Human Clinical

"Sequencing of human tumors has, however, identified additional secondary mutations in approximately 50% of EHE, most commonly the loss of tumor suppressor CDKN2A."

Establishes CDKN2A loss as the most frequent secondary alteration in human EHE. Note that this statement summarises prior sequencing studies in the paper's Purpose section rather than reporting this paper's own cohort.

PMID:31537895 SUPPORT Human Clinical

"Memorial Sloan Kettering-IMPACT targeted DNA sequencing was performed to identify secondary genetic alterations showing more than half of tumors had a genetic alteration beyond the disease-defining gene fusion."

Primary sequencing cohort corroborating that secondary alterations occur in more than half of EHE tumours.

PMID:36598859 SUPPORT Model Organism

"Loss of Cdkn2a within EHE was associated with more aggressive disease, as displayed by earlier tumor-related morbidity/mortality and enhanced tumor cell proliferation."

Functional mouse evidence that CDKN2A loss accelerates EHE progression, supporting its role as a modifier node and its route through enhanced proliferation.

Multifocal Dissemination and Serosal Involvement

EHE is characteristically multifocal, and multifocality is best interpreted as early metastatic colonisation rather than independent primary tumours. Successful colonisation of lung, liver, bone, lymph nodes and - most ominously - pleural and peritoneal surfaces determines outcome. Pleural involvement and lymph node metastasis confer an aggressive course akin to high-grade sarcoma, and serosal effusion is the single most consistent adverse prognostic feature.

cell population proliferation GO:0008283 ↑ INCREASED cell migration GO:0016477 ↑ INCREASED

pleura UBERON:0000977

Show evidence (4 references)

PMID:31537895 SUPPORT Human Clinical

"Multifocality, pleural involvement, lymph node or distant metastases had a significantly worse outcome."

Establishes multifocal and metastatic colonisation, especially pleural, as the outcome-determining step in EHE.

PMID:37444414 SUPPORT Human Clinical

"No survival differences were observed between multifocal and metastatic disease, suggesting that multifocality represents early metastases"

Clinical support for interpreting EHE multifocality as metastatic colonisation.

PMID:31537895 SUPPORT Human Clinical

"multifocal liver epithelioid hemangioendothelioma represents a clonal process, demonstrating that separate lesions harbor fusions with an identical genomic breakpoint, in keeping with a single clone, i.e. metastatic dissemination confined to one organ"

Direct molecular (shared-breakpoint) evidence that multifocal hepatic EHE is clonal metastatic dissemination rather than independent multicentric transformation. This is the strongest basis for conformance to invasion_and_metastasis#Metastatic Colonization.

+ 1 more reference

✶

Histopathology

4

Epithelioid Cells in Myxohyaline Stroma

Conventional EHE shows cords, strands and single files of epithelioid endothelial cells set in a distinctive myxoid, myxochondroid or sclerotic (hyaline) extracellular stroma, generally without well-formed vascular channels.

Show evidence (1 reference)

PMID:32913529 SUPPORT Human Clinical

"Histologically, EHE typically showed a mucous hyaline or cartilaginous stroma, often forming a primitive vascular lumen, and expressed vascular endothelial markers."

Directly documents the myxohyaline/cartilaginous stroma as the typical histologic background of EHE.

Intracytoplasmic Vascular Lumina (Blister Cells)

Rather than forming mature vascular channels, conventional EHE cells contain immature intracytoplasmic lumina or vacuoles ("blister cells"), which represent primitive vascular differentiation and are a key discriminator from epithelioid haemangioma and epithelioid angiosarcoma.

Show evidence (1 reference)

PMID:21584898 SUPPORT Human Clinical

"the most important histologic criteria applied to distinguish EHE from both EH and E-AS was the lack of well-formed vascular channel formation in EHE, with only immature, intra-cytoplasmic lumina being observed"

Documents intracytoplasmic lumina as the defining vasoformative feature distinguishing EHE from its principal mimics.

Atypical Histology (Size and Mitotic Activity)

A subset of EHE shows atypical histology, defined as at least two of high mitotic activity, high nuclear grade and coagulative necrosis. Together with tumour size greater than 30 mm this defines a validated three-tier risk stratification.

Show evidence (1 reference)

PMID:33729740 SUPPORT Human Clinical

"Twenty-two tumors (35.5%) demonstrated atypical histology, defined by having at least 2 of the following 3 findings: high mitotic activity (>1/2 mm2), high nuclear grade, and coagulative necrosis."

Defines the atypical-histology criteria and their frequency in a molecularly confirmed cohort.

Vasoformative Architecture with Eosinophilic Cytoplasm

YAP1-TFE3-fused tumours differ morphologically from conventional EHE: solid sheets of coalescing nests, a pseudoalveolar/(pseudo)vasoformative pattern, and discohesive strands in myxohyaline stroma, with large epithelioid cells having abundant glassy eosinophilic to vacuolated cytoplasm.

Show evidence (1 reference)

PMID:34381186 SUPPORT Human Clinical

"we describe three dominant architectural patterns: solid sheets of coalescing nests, pseudoalveolar and (pseudo)vasoformative pattern, and discohesive strands and clusters of cells set in a myxoid to myxohyaline stroma."

Characterises the distinct architecture of the YAP1-TFE3 subtype.

⬡

Pathograph

Use the checkboxes to hide or show graph categories. Hover nodes for evidence and cross-linked metadata.

●

Phenotypes

10

Blood 1

Anemia HP:0001903

Show evidence (1 reference)

PMID:33107994 SUPPORT Other

"pulmonary disease, weight loss, anemia, and pulmonary hemorrhage were correlated with poorer patient outcomes"

Names anaemia as an EHE feature correlated with poorer outcome. Evidence source is OTHER because this is an editorial commentary summarising the cohort literature.

Digestive 2

Hepatic Involvement Neoplasm of the liver HP:0002896

Show evidence (1 reference)

PMID:37260142 SUPPORT Human Clinical

"Skin and connective tissues were the most common presenting sites (33.4%), followed by liver (24.5%), lung (17.6%), and bone (12.5%)."

Establishes the liver as a major site of disease. Support is PARTIAL and no `frequency` band is asserted because this figure is the distribution of *presenting primary site* in the broader malignant-hemangioendothelioma registry group, not the fraction of EHE patients with hepatic involvement - which is higher, since most patients are multifocal.

Ascites HP:0001541

Show evidence (2 references)

PMID:33107994 SUPPORT Other

"It has been demonstrated that patients with effusions, either pleural or peritoneal, have worse outcomes."

Names peritoneal effusion explicitly as an adverse EHE feature. Evidence source is OTHER because this is an editorial commentary summarising the cohort literature.

PMID:33107985 SUPPORT Human Clinical

"at baseline, 13 of 38 patients had the presence of serosal effusions and systemic symptoms"

Documents serosal effusions in roughly a third of patients with advanced EHE. Support is PARTIAL because the report aggregates pleural and peritoneal effusion rather than reporting ascites separately.

Metabolism 1

Pleural Effusion HP:0002202

Show evidence (1 reference)

PMID:33107985 SUPPORT Human Clinical

"Serosal effusions are confirmed as an unfavorable prognostic sign associated with short survival, and sirolimus displays limited activity in this subgroup."

Supports serosal effusion as both a clinical manifestation and a treatment-resistant adverse prognostic feature.

Growth 1

Systemic Symptoms Weight loss HP:0001824

Show evidence (2 references)

PMID:33107994 SUPPORT Other

"pulmonary disease, weight loss, anemia, and pulmonary hemorrhage were correlated with poorer patient outcomes"

Names weight loss explicitly as an EHE feature correlated with poorer outcome. Evidence source is OTHER because this is an editorial commentary summarising the cohort literature.

PMID:33107985 SUPPORT Human Clinical

"at baseline, 13 of 38 patients had the presence of serosal effusions and systemic symptoms"

Documents systemic symptoms as a recognised baseline feature of advanced EHE. Support is PARTIAL because the source does not itemise weight loss separately from other constitutional symptoms.

Neoplasm 2

Pulmonary Involvement FREQUENT Neoplasm of the lung HP:0100526

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"In combination with other metastatic foci, such as the liver or bone, 49 cases (59%) involved the lung at diagnosis."

In a cohort of 83 molecularly confirmed WWTR1-CAMTA1-positive EHE patients, 59% had lung involvement at diagnosis, mapping to the FREQUENT (30-79%) band. This directly measures organ involvement in EHE, unlike registry primary-site distributions.

Soft Tissue Involvement OCCASIONAL Soft tissue neoplasm HP:0031459

Show evidence (2 references)

PMID:31537895 SUPPORT Human Clinical

"A total of 17 patients (20% of the whole WWTR1-CAMTA1-positive cohort) had primary soft tissue tumors, all but one had solitary disease."

Quantifies primary soft tissue disease at 20% of a molecularly confirmed EHE cohort, mapping to the OCCASIONAL (5-29%) band, and notes its characteristically solitary presentation.

PMID:31537895 SUPPORT Human Clinical

"Soft tissue epithelioid hemangioendothelioma were frequently solitary, followed an uneventful clinical course being often managed with curative surgery."

Supports the favourable, surgically curable behaviour of soft tissue EHE.

Other 3

Vascular Sarcoma HP:0100242

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"Epithelioid hemangioendothelioma is a low-grade malignant vascular tumor with an intermediate clinical behavior between benign hemangiomas and high-grade angiosarcomas."

Defines the core disease identity as a low-grade malignant vascular neoplasm.

Skeletal Involvement Neoplasm of the skeletal system HP:0010622

Show evidence (1 reference)

PMID:37260142 SUPPORT Human Clinical

"Skin and connective tissues were the most common presenting sites (33.4%), followed by liver (24.5%), lung (17.6%), and bone (12.5%)."

Establishes bone as a recognised site of disease. Support is PARTIAL and no `frequency` band is asserted because this figure is the distribution of presenting primary site in the broader malignant-hemangioendothelioma registry group, not the fraction of EHE patients with skeletal involvement.

Metastatic Disease FREQUENT

No `term:` binding is asserted. HPO has no generic metastasis term (HP:0200059 is Metastatic angiosarcoma, a different disease) and the available generic HP:0002664 Neoplasm would encode nothing beyond "this is a cancer"; metastasis is better represented from NCIT/SNOMED, which is not in scope for `phenotype_term`.

Show evidence (2 references)

PMID:31537895 SUPPORT Human Clinical

"The majority of patients (n=61; 73%) had multifocal primary disease at the time of diagnosis and 70% (n=58) developed distant metastases."

In a molecularly confirmed EHE cohort, 70% developed distant metastases and 73% were multifocal at diagnosis, mapping to the FREQUENT (30-79%) band.

PMID:37260142 SUPPORT Human Clinical

"At presentation, 36.4% of patients had localized disease; 21.6% presented with regional and 41.7% with distant metastases."

Population-registry corroboration that a plurality of patients already have distant spread at diagnosis.

🧬

Genetic Associations

5

WWTR1-CAMTA1 (Somatic t(1;3)(p36.23;q25.1) gene fusion)

Gene: WWTR1 hgnc:24042

Show evidence (2 references)

PMID:21885404 SUPPORT Human Clinical

"The gene fusion is present in virtually all EHEs tested but is absent from all other vascular neoplasms, demonstrating it to be a disease-defining genetic alteration."

Establishes the fusion as necessary and specific for EHE among vascular neoplasms.

PMID:35740643 SUPPORT Other

"Herein, we present a unique TAZ-driven cancer, epithelioid hemangioendothelioma (EHE), which harbors a WWTR1(TAZ)-CAMTA1 gene fusion in at least 90% of cases."

Review-level source for the approximately 90% prevalence of the fusion. Evidence source is OTHER because this is a review article.

CAMTA1 (Somatic gene fusion partner (3-prime partner of WWTR1-CAMTA1))

Gene: CAMTA1 hgnc:18806

Show evidence (1 reference)

PMID:21885404 SUPPORT Human Clinical

"Thus, CAMTA1 expression is activated inappropriately through a promoter-switch mechanism."

Establishes the mechanism by which the CAMTA1 partner is misexpressed in endothelium.

YAP1 (Somatic gene fusion (YAP1-TFE3) in a minority subset)

Gene: YAP1 hgnc:16262

Show evidence (3 references)

PMID:26414223 SUPPORT Human Clinical

"resulting in the formation of a WWTR1-CAMTA1 fusion gene in approximately 90% of cases; a small subset (<5%) have a YAP1-TFE3 fusion gene."

Supports the lower bound of the range (<5%) alongside the ~90% WWTR1-CAMTA1 figure.

PMID:32913529 SUPPORT Human Clinical

"Approximately 90% and 5% of EHE cases are associated with the WWTR1-CAMTA1 and YAP1/TFE3 fusion gene, respectively."

Independent source giving the central estimate of about 5%.

PMID:33766984 SUPPORT Other

"YAP1-TFE3 (in ∼10% of cases)"

Supports the upper bound of the reported range. Evidence source is OTHER because this is a background statement in the introduction of a mechanistic paper rather than its own cohort measurement.

TFE3 (Somatic gene fusion partner (3-prime partner of YAP1-TFE3))

Gene: TFE3 hgnc:11752

Show evidence (1 reference)

PMID:23737213 SUPPORT Human Clinical

"All tumors expressed endothelial markers, as well as strong nuclear TFE3."

Confirms TFE3 protein misexpression as the functional consequence of the fusion.

CDKN2A (Secondary somatic loss)

Gene: CDKN2A hgnc:1787

Show evidence (1 reference)

PMID:36598859 SUPPORT Human Clinical

"Sequencing of human tumors has, however, identified additional secondary mutations in approximately 50% of EHE, most commonly the loss of tumor suppressor CDKN2A."

Establishes the frequency and identity of the dominant secondary alteration.

💊

Medical Actions

11

Complete Surgical Resection

Action: Definitive Surgical Resection NCIT:C154430

Complete resection is the mainstay and the only reliably curative option for localised, resectable EHE, and should be planned at a sarcoma reference centre. There is no established routine adjuvant systemic therapy or radiotherapy.

Show evidence (1 reference)

PMID:34381186 SUPPORT Human Clinical

"Most patients (88%) were treated with primary surgical resection."

Documents primary surgical resection as the dominant real-world management approach in a genetically confirmed cohort.

Active Surveillance

Because a substantial proportion of patients have stable, asymptomatic disease for years, an initial period of observation is often preferred to avoid the toxicity of systemic therapy. Objective progression, symptomatic deterioration, anaemia or new serosal involvement should prompt reconsideration of systemic treatment.

Show evidence (2 references)

PMID:29546487 SUPPORT Other

"Our treatment strategy is watch-and-wait versus active therapy on clinical trial or based on results of prior clinical trials."

Expert-review statement of the watch-and-wait strategy. Evidence source is OTHER because this is a narrative treatment review.

PMID:31537895 SUPPORT Human Clinical

"One-third of these patients (n=18) were managed initially with expectant observation"

Quantifies real-world use of initial observation in a molecularly confirmed EHE cohort, showing this is a mainstream rather than exceptional approach.

Sirolimus

Action: Pharmacotherapy NCIT:C15986

Agent: sirolimus CHEBI:9168

mTOR inhibition has the strongest recurring disease-specific activity signal in EHE. In a 38-patient retrospective series of progressive disease, sirolimus produced partial response in 10.8% and stable disease in 75.7%, with median progression-free survival of 13 months. Activity is markedly reduced in patients with serosal effusions. Use is off-label in most jurisdictions.

Mechanism Target:

INHIBITS mTORC1 Pathway Activation — Sirolimus inhibits mTOR complex 1, the node on which fusion-driven YAP/TAZ activity converges.

Show evidence (3 references)

PMID:33107985 SUPPORT Human Clinical

"The best RECIST responses were a partial response in 4 patients (10.8%), stable disease in 28 patients (75.7%), and disease progression in 5 patients (13.5%)."

Quantifies the activity of sirolimus in progressive EHE.

PMID:33107994 SUPPORT Other

"The results from a robust retrospective cohort study support the use of sirolimus for patients who have multifocal epithelioid hemangioendothelioma without serosal effusion."

Independent editorial appraisal endorsing sirolimus in effusion-negative multifocal EHE. Evidence source is OTHER because this is a commentary.

PMID:39283723 SUPPORT In Vitro

"Sirolimus exhibited markedly higher antitumor activity compared with doxorubicin, concurrently reducing GDF-15 expression/release both in vivo and in vitro."

Preclinical patient-derived model supporting sirolimus over anthracycline chemotherapy in EHE.

Trametinib

Action: Targeted Therapy NCIT:C93352

Agent: trametinib CHEBI:75998

MEK inhibition is the mechanistically rationalised targeted approach, derived from the demonstration that TAZ-CAMTA1 signals through CTGF to the RAS-MAPK cascade. Trametinib abrogated fusion-driven anchorage-independent growth and xenograft expansion, which supported the SARC033 phase II trial in unresectable or metastatic EHE.

Mechanism Target:

INHIBITS CTGF-Integrin-RAS-MAPK Signaling — Trametinib inhibits MEK downstream of the fusion-driven CTGF-RAS cascade, the node that sustains anchorage-independent growth.

Show evidence (2 references)

PMID:35443056 SUPPORT In Vitro

"Further, pharmacologic inhibition of MAPK signaling through PD 0325901 and trametinib abrogated TC-driven anchorage-independent growth."

Preclinical basis for MEK inhibition as a fusion-directed therapy in EHE.

PMID:35443056 SUPPORT In Vitro

"These studies provided the preclinical rationale for SARC033 (NCI 10015-NCT03148275), a nonrandomized, open-label, phase II study of trametinib in patients with unresectable or metastatic EHE."

Links the preclinical mechanism directly to the clinical trial testing it.

Investigational TAZ-TEAD Directed Therapy

Action: Targeted Therapy NCIT:C93352

Because the fusion oncoprotein acts through TEAD and established tumours regress when it is switched off, direct disruption of the TAZ-CAMTA1-TEAD interaction is the most rational fusion-directed strategy. TEAD inhibitors have been tested against the first EHE cell lines, and a mechanistically distinct approach - restoring cytoplasmic sequestration of the fusion with a statin - abrogates fusion-driven anchorage-independent growth in endothelial models. Both remain preclinical/investigational with no clinical efficacy data in EHE.

Mechanism Target:

INHIBITS Constitutive Nuclear TAZ-TEAD Transcriptional Activation — TEAD inhibition blocks the fusion's transcriptional output; statins act upstream by promoting its cytoplasmic sequestration.

Show evidence (2 references)

PMID:36598859 SUPPORT In Vitro

"EHE cell lines were established through ex vivo culture of tumor cells and evaluated for sensitivity to TEAD inhibition and trametinib."

Documents TEAD inhibition as an actively tested fusion-directed strategy in the first EHE cell lines.

PMID:33766984 SUPPORT In Vitro

"Therefore, therapies that are known to inhibit YAP/TAZ activity also may inhibit TAZ-CAMTA1-induced endothelial cell transformation."

Supports the broader class rationale, demonstrated with simvastatin abrogating soft-agar growth of TAZ-CAMTA1-expressing endothelial cells.

Liver Transplantation

Action: Liver Transplantation NCIT:C15271

Selected patients with unresectable but liver-confined hepatic EHE may be considered for transplantation at specialist centres. Selection is highly specialised and the supporting evidence is registry- and series-based.

Conventional Cytotoxic Chemotherapy

Action: Chemotherapy NCIT:C15632

Agent: doxorubicin CHEBI:28748 gemcitabine CHEBI:175901 paclitaxel CHEBI:45863

Single- and multiagent doxorubicin-, gemcitabine-, and taxane-based regimens are used for systemic treatment of EHE, but there is no established standard cytotoxic regimen and the cited observational cohort does not establish controlled efficacy.

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"Both single and multiagent cytotoxic regmens were also commonly used, either doxorubicin-, gemcitabine-, or taxane-based."

Documents real-world use of these cytotoxic regimen classes. Support is PARTIAL because the observational cohort does not provide controlled evidence of efficacy.

Local-Control Radiation Therapy

Action: Radiation Therapy NCIT:C15313

Radiation may be used as a local-control procedure alongside systemic treatment in selected EHE patients. This is distinct from routine adjuvant radiotherapy, which is not established after complete resection.

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"local treatment modalities, such as surgery, embolization, or radiation, were frequently utilized in conjunction with systemic therapy and often contributed to prolonged disease control."

Identifies radiation among local-control modalities used with systemic therapy. Support is PARTIAL because this observational statement does not isolate radiation's effect or provide controlled efficacy evidence.

Local-Control Embolization

Action: Embolization Therapy NCIT:C15230

Embolization may be used as a local-control procedure alongside systemic treatment in selected EHE patients.

Show evidence (1 reference)

PMID:31537895 SUPPORT Human Clinical

"local treatment modalities, such as surgery, embolization, or radiation, were frequently utilized in conjunction with systemic therapy and often contributed to prolonged disease control."

Identifies embolization among local-control modalities used with systemic therapy. Support is PARTIAL because this observational statement does not isolate embolization's effect or provide controlled efficacy evidence.

Paclitaxel plus Bevacizumab

Action: Targeted Therapy NCIT:C93352

Agent: bevacizumab NCIT:C2039 paclitaxel CHEBI:45863

Paclitaxel plus the anti-VEGF monoclonal antibody bevacizumab produced partial responses in a small hepatic EHE series, with prolonged stabilisation during bevacizumab maintenance in one patient. This is not an established standard of care.

Show evidence (1 reference)

PMID:32201052 SUPPORT Human Clinical

"While combination of paclitaxel and bevacizumab resulted in partial response in 3 patients, one of them also achieved long-term disease stabilization with bevacizumab maintenance with no adverse event."

Reports the actual observed responses to VEGF-directed therapy. Support is PARTIAL because this is a four-patient case series, not a controlled trial.

Pazopanib

Action: Targeted Therapy NCIT:C93352

Agent: pazopanib CHEBI:71219

The oral multikinase inhibitor pazopanib produced clinical benefit in two patients in a small hepatic EHE series. It is not an established standard of care, and the supporting evidence is case-series level only.

Show evidence (1 reference)

PMID:32201052 SUPPORT Human Clinical

"Two of the patients had clear benefit from pazopanib during the course of disease."

Supports pazopanib activity in hepatic EHE at case-series level.

🔬

Biochemical Markers

4

Nuclear CAMTA1 Immunohistochemistry

Show evidence (2 references)

PMID:26414223 SUPPORT Human Clinical

"In total, 51/59 cases (86%) of EHE showed diffuse nuclear staining for CAMTA1, including 44/48 cases (92%) with conventional histology and 7/11 cases (64%) with "malignant" histology."

Quantifies the sensitivity of nuclear CAMTA1 immunohistochemistry for EHE.

PMID:32913529 SUPPORT Human Clinical

"Twelve of the 15 EHE samples showed positive nuclear CAMTA1 expression with immunohistochemistry, whereas six of the 37 histologic mimics showed positive nuclear expression."

Tempers the claim: CAMTA1 immunohistochemistry is useful but imperfectly specific, which is why molecular fusion confirmation is preferred.

WWTR1-CAMTA1 Fusion Detection by FISH or RNA Sequencing

Show evidence (1 reference)

PMID:32913529 SUPPORT Human Clinical

"FISH detected a red-green signal fusion in 14 of the 15 cases of EHE, but in none of the 10 vascular tumors."

Quantifies FISH sensitivity and specificity for the defining fusion.

Endothelial Marker Immunohistochemistry (CD31, ERG, CD34)

Show evidence (1 reference)

PMID:34381186 SUPPORT Human Clinical

"Immunohistochemically, ERG, CD31, and TFE3 were consistently expressed, while expression of CD34 (83%) and cytokeratin AE1/AE3 (20%) was variable."

Documents the endothelial immunophenotype and the variable cytokeratin expression that can cause diagnostic confusion.

Circulating GDF-15

Show evidence (2 references)

PMID:39283723 SUPPORT Human Clinical

"Blood sample analyses from two independent patient series showed a significant correlation between GDF-15 and EHE aggressiveness."

Human evidence linking circulating GDF-15 to EHE aggressiveness.

PMID:39283723 SUPPORT In Vitro

"This reduction was attributed to the drug-induced inhibition of phosphorylation/activation of 4E-BP1 and subsequent downregulation of the GDF-15 transcription factors ATF4 and ATF5."

Mechanistic basis for GDF-15 as a pharmacodynamic readout of mTOR inhibition in EHE.

📈

Progression

3

Heterogeneous course from indolent to rapidly fatal

EHE has a markedly heterogeneous natural history. Solitary soft-tissue disease is often managed with curative surgery and follows an uneventful course, while pleural involvement or lymph node metastasis produces a course akin to high-grade sarcoma. Five-year overall survival was 59% in the WWTR1-CAMTA1 subset versus 86% in the YAP1-TFE3 subset in a large single-institution translocation-positive cohort, and only 22% and 30% in patients with pleural disease or lymph node metastases respectively.

Show evidence (2 references)

PMID:31537895 SUPPORT Human Clinical

"Patients with conventional epithelioid hemangioendothelioma with WWTR1-CAMTA1 fusion had a less favorable outcome compared with the YAP1-TFE3 subset, the 5-year overall survival being 59% versus 86%, respectively."

Quantifies the survival difference between the two molecular subsets.

PMID:31537895 SUPPORT Human Clinical

"Patients with pleural disease or lymph node metastases had an aggressive clinical course akin to high-grade sarcomas, with 22% and 30%, respectively, alive at 5 years, compared with >70% survival rate in patients lacking these two adverse factors."

Quantifies the prognostic impact of serosal and nodal involvement.

Risk stratification by tumour size and histologic atypia

A proposed three-tier system combining tumour size greater than 30 mm with atypical histology separates low-, intermediate- and high-risk groups with 5-year overall survival of 100%, 81.8% and 16.9%. It has since been applied successfully to unifocal disease in an independent nationwide cohort.

Show evidence (2 references)

PMID:37444414 SUPPORT Human Clinical

"In unifocal tumors, survival could be predicted using the risk stratification model of Shibayama et al., dividing the cases into low- (n = 4), intermediate- (n = 15) and high- (n = 3) risk groups."

Independent nationwide cohort applying and supporting the proposed risk-stratification model in unifocal disease.

PMID:33729740 SUPPORT Human Clinical

"A proposed 3-tiered risk assessment system using these 2 parameters significantly stratified the patients into low-risk, intermediate-risk, and high-risk groups with 5-year overall survival rates of 100%, 81.8%, and 16.9%, respectively (P<0.001)."

Provides the validated risk-stratification model and its survival separation.

Stage-dependent survival at presentation

Extent of disease at diagnosis strongly predicts outcome, and most deaths occur early in the disease course.

Show evidence (1 reference)

PMID:37260142 SUPPORT Human Clinical

"Observed survival at 3 years was 79.7%, 70.7%, and 46.0% for patients presenting with local, regional, and distant disease and most deaths occurred within the first 2 years."

Population-level survival by extent of disease at presentation.

📊

Prevalence

1

United States

Annual Incidence 0.04 per 100,000 <1 in 1,000,000

US Cancer Statistics / SEER combined database, 2001-2017: 0.4 cases per million person-years for malignant hemangioendothelioma overall, the majority of which are epithelioid. Normalised to 0.04 per 100,000.

Show evidence (1 reference)

PMID:37260142 SUPPORT Human Clinical

"The incidence of hemangioendothelioma in the US is 0.4 cases per million person-years."

Population-registry incidence estimate for the hemangioendothelioma group in which EHE predominates.

🔀

Differential Diagnoses

4

Conditions with similar clinical presentations that must be differentiated from Epithelioid Hemangioendothelioma:

Epithelioid angiosarcoma

Overlapping Features The principal high-grade mimic. No `disease_term` is bound here because MONDO currently has no separate class for epithelioid angiosarcoma - it is listed as an EXACT synonym of MONDO:0015523 itself (inherited from NCIT:C3800). That is an ontology defect: the two are distinct entities with different genetics, treatment and prognosis, and binding the EHE term to its own differential would propagate the error.

Distinguishing Features

Marked nuclear pleomorphism, brisk mitotic activity and necrosis
Negative for both WWTR1-CAMTA1 and YAP1-TFE3 fusions
High-grade clinical behaviour rather than intermediate malignancy

Show evidence (1 reference)

PMID:21584898 SUPPORT Human Clinical

"This recurrent translocation has not been detected in any of the morphologic mimics of EHE, such as EH, E-AS or epithelioid sarcoma-like EHE, and thus can serve as a useful molecular diagnostic tool in challenging cases."

Demonstrates that the defining fusion is absent from epithelioid angiosarcoma, providing the molecular discriminator.

Epithelioid hemangioma Not Yet Curated MONDO:0021169

Overlapping Features The principal benign mimic. Misdiagnosis of epithelioid haemangioma as EHE leads to overtreatment.

Distinguishing Features

Benign, lobulated or well-circumscribed lesion
Clearly vasoformative with mature open vascular lumina
Bland cytomorphology and absence of CAMTA1/WWTR1 rearrangement

Show evidence (1 reference)

PMID:21584898 SUPPORT Human Clinical

"Particularly important is the lack of CAMTA1 and WWTR1 rearrangements in EH, an under-recognized benign tumor, that is often mis-diagnosed as EHE, resulting in overtreatment."

Directly documents epithelioid haemangioma as the key benign mimic and the clinical cost of confusing it with EHE.

Pseudomyogenic (epithelioid sarcoma-like) hemangioendothelioma Not Yet Curated MONDO:0975754

Distinguishing Features

Rhabdomyoblast-like cells with dense eosinophilic cytoplasm
Co-expression of endothelial and epithelial markers
Typically multiple superficial nodules; negative for the EHE-defining fusions

Show evidence (1 reference)

PMID:23737213 SUPPORT Human Clinical

"A control group, including 18 epithelioid hemangiomas, nine pseudomyogenic HE, and three epithelioid angiosarcomas, was also tested."

Confirms pseudomyogenic haemangioendothelioma as a routine comparator that was fusion-negative in the TFE3 screen. Support is PARTIAL because the abstract does not itemise its alternative driver.

Alveolar soft part sarcoma MONDO:0011655

Distinguishing Features

Also strongly nuclear TFE3-positive with a pseudoalveolar pattern
Negative for the endothelial markers CD31 and ERG
Carries an ASPSCR1-TFE3 rather than YAP1-TFE3 fusion

Show evidence (1 reference)

PMID:23737213 SUPPORT Human Clinical

"The presence of diffuse expression of CD31 and/or ERG endothelial markers helps in the distinction from other TFE3-positive neoplasms, such as alveolar soft part sarcoma, PEComa, and Xp11-translocation positive renal cell carcinomas."

Gives the immunophenotypic discriminator between YAP1-TFE3 EHE and other TFE3-driven tumours.

🔬

Clinical Trials

2

NCT03148275 PHASE_II COMPLETED

SARC033 (NCI 10015): non-randomised, open-label phase II study of the MEK inhibitor trametinib in patients with unresectable, locally advanced or metastatic epithelioid hemangioendothelioma. Built directly on the CTGF-RAS-MAPK dependence demonstrated preclinically.

Show evidence (1 reference)

clinicaltrials:NCT03148275 SUPPORT Human Clinical

"This phase II trial studies how well trametinib works in treating patients with epithelioid hemangioendothelioma that has spread to other places in the body (metastatic), nearby tissue or lymph nodes (locally advanced), or cannot be removed by surgery (unresectable)."

Confirms the trial design and population for the fusion-directed MEK inhibition strategy.

NCT06408441 RECRUITING

The EURACAN prospective clinical registry dedicated to EHE - a registry-based cohort study of newly diagnosed, molecularly confirmed EHE across 21 sarcoma reference centres in 10 European countries, designed to resolve the natural-history, prognostic-factor and treatment-efficacy gaps that dominate this disease.

Show evidence (1 reference)

clinicaltrials:NCT06408441 SUPPORT Human Clinical

"The objectives are to improve the understanding of EHE natural history, validate and identify new prognostic and predictive factors, clarify the activity and efficacy of currently available treatment options, describe treatment pattern."

Documents the registry addressing the principal evidence gaps in EHE.

🧫

Experimental Models

3

Conditional Wwtr1-Camta1 Knock-In Mouse

The fusion is targeted to the endogenous Wwtr1 locus so that it is controlled by native transcriptional regulators upon Cre activation. Mice develop tumours indistinguishable from human EHE and no other tumour type, demonstrating both sufficiency and exquisite tissue specificity.

Show evidence (1 reference)

PMID:33766982 SUPPORT Model Organism

"Overall, these results demonstrate unequivocally that TAZ-CAMTA1 is sufficient to drive EHE formation with exquisite specificity, as no other tumor types were observed."

Establishes the model's fidelity and the sufficiency of the fusion.

Endothelial-Specific TAZ-CAMTA1 Mouse

Expression of TAZ-CAMTA1 restricted to endothelial cells generates vascular tumours with the distinctive features of EHE; inhibiting the fusion causes regression, establishing continuing oncogene dependence.

Show evidence (1 reference)

PMID:33766984 SUPPORT Model Organism

"Our study provides the first genetic model of a TAZ fusion oncoprotein driving its associated human cancer, pinpointing TAZ-CAMTA1 as the key driver and a valid therapeutic target of EHE."

Confirms the endothelial cell of origin and validates the fusion as a therapeutic target.

EHE Cell Lines and Patient-Derived Xenograft

The first EHE cell lines were generated from Cdkn2a-null TAZ-CAMTA1 mouse tumours and are addicted to the fusion oncoprotein; a separate 2024 patient-derived xenograft with a matched cell line retains the WWTR1::CAMTA1 fusion and the donor tumour's histomorphology and genomic/transcriptomic profile.

Show evidence (2 references)

PMID:36598859 SUPPORT In Vitro

"These cell lines are "addicted" to the TC oncoprotein, replicate the EHE transcriptional profile, and generate EHE tumors when injected into immunodeficient mice."

Documents oncogene addiction in the first EHE cell lines, the cellular basis for fusion-directed therapy.

PMID:39283723 SUPPORT In Vitro

"A patient-derived xenograft model and corresponding cell line were established from a patient with advanced EHE, demonstrating consistency with the original tumor in terms of histomorphology, WWTR1::CAMTA1 fusion presence, and genomic and transcriptomic profiles."

Documents the human patient-derived model used for therapeutic and biomarker testing.

{ }

Source YAML

click to show

name: Epithelioid Hemangioendothelioma
creation_date: "2026-08-01T00:00:00Z"
category: Cancer
categories:
- Sarcoma
- Soft Tissue Sarcoma
- Vascular Tumor
- Rare Cancer
parents:
- vascular cancer
- hemangioendothelioma
disease_term:
  preferred_term: epithelioid hemangioendothelioma
  term:
    id: MONDO:0015523
    label: epithelioid hemangioendothelioma
description: >-
  Epithelioid hemangioendothelioma (EHE) is an ultra-rare, low-grade malignant
  vascular sarcoma of endothelial lineage whose clinical behaviour sits between
  benign haemangioma and high-grade angiosarcoma. It is a fusion-defined
  entity: a somatic t(1;3)(p36;q25) translocation generates a
  WWTR1(TAZ)-CAMTA1 chimeric transcriptional coactivator in roughly 90% of
  cases, and a smaller subset instead carries a YAP1-TFE3 fusion. Both fusions
  short-circuit the Hippo pathway, producing a constitutively nuclear TEAD
  coactivator that reprograms endothelial cells. EHE most often involves liver,
  lung, bone and soft tissue, is frequently multifocal at presentation, and has
  a strikingly heterogeneous course ranging from years of radiographic
  stability to rapidly fatal dissemination.
synonyms:
- EHE
- epithelioid haemangioendothelioma
- epithelioid angioendothelioma
- malignant epithelioid hemangioendothelioma
- intravascular bronchioloalveolar tumor
notes: >-
  Scope note: this entry models conventional WWTR1-CAMTA1 epithelioid
  hemangioendothelioma together with the YAP1-TFE3 variant, which is captured
  as a subtype. Some contemporary expert groups argue that YAP1-TFE3-fused
  hemangioendothelioma should be split out as a distinct entity on
  morphologic, immunophenotypic and prognostic grounds; MONDO:0015523 does not
  yet make that split, so the two are kept together here with the
  differentiating features recorded on the subtype. Note also that the MONDO
  record carries "epithelioid angiosarcoma" as an EXACT synonym (inherited from
  NCIT:C3800); this is a known ontology defect - epithelioid angiosarcoma is a
  mechanistically and prognostically distinct high-grade tumour that is
  WWTR1-CAMTA1-negative and must not be conflated with EHE.

has_subtypes:
- name: WWTR1-CAMTA1
  display_name: Conventional EHE (WWTR1-CAMTA1 fusion)
  subtype_term:
    preferred_term: EHE with WWTR1-CAMTA1 gene fusion
    term:
      id: NCIT:C177552
      label: Epithelioid Hemangioendothelioma with WWTR1-CAMTA1 Gene Fusion
  description: >-
    The conventional and by far the most common form, accounting for
    approximately 90% of EHE. Defined by the t(1;3)(p36.23;q25.1)
    WWTR1(TAZ)-CAMTA1 fusion. Morphology is cords, strands and nests of
    epithelioid endothelial cells in a myxohyaline stroma, typically without
    well-formed vascular channels. Nuclear CAMTA1 immunohistochemistry is the
    surrogate marker.
- name: YAP1-TFE3
  display_name: YAP1-TFE3-fused hemangioendothelioma
  subtype_term:
    preferred_term: EHE with YAP1-TFE3 gene fusion
    term:
      id: NCIT:C177553
      label: Epithelioid Hemangioendothelioma with YAP1-TFE3 Gene Fusion
  description: >-
    A minority molecular subset (about 5% of cases, reported from <5% up to
    ~10%) occurring in younger adults, with more overt vasoformative
    architecture, voluminous eosinophilic to vacuolated cytoplasm, strong
    nuclear TFE3 expression and negative CAMTA1. Reported survival is higher
    than in the WWTR1-CAMTA1 subset (86% vs 58-59% at 5 years) despite a high
    propensity for multifocal and metastatic presentation, but the difference
    did not reach statistical significance in the largest comparison
    (p = 0.095), so the prognostic separation should be treated as suggestive
    rather than established.
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Although the disease specific survival rate at 5 years was 86% in
      TFE3-rearranged patients compared to 58% in CAMTA1-rearranged patients,
      the difference between groups did not reach statistical signficance
      (p = 0.095).
    explanation: >-
      Gives both the survival difference and the explicit statement that it was
      not statistically significant, which is why the subtype prognostic claim
      is hedged. Support is PARTIAL for this reason.

pathophysiology:
- name: WWTR1-CAMTA1 Fusion Oncogene Formation
  biological_scale: MOLECULAR
  role: trigger
  description: >-
    A somatic t(1;3)(p36.23;q25.1) translocation fuses the 5' portion of WWTR1
    (encoding the Hippo effector TAZ) on 3q25 to the 3' portion of CAMTA1 on
    1p36. The chimeric gene is driven by the WWTR1 promoter, which is highly
    active in endothelial cells, so CAMTA1 sequence - normally expressed only
    in brain - is inappropriately activated by a promoter-switch mechanism. The
    fusion retains the TAZ 14-3-3-binding and WW domains together with the
    CAMTA1 TIG DNA-binding domain, ankyrin repeats and IQ calmodulin-binding
    motifs.
  genes:
  - preferred_term: WWTR1
    term:
      id: hgnc:24042
      label: WWTR1
  - preferred_term: CAMTA1
    term:
      id: hgnc:18806
      label: CAMTA1
  gene_products:
  - preferred_term: TAZ-CAMTA1 fusion oncoprotein
    term:
      id: NCIT:C121673
      label: WWTR1/CAMTA1 Fusion Protein
  cell_types:
  - preferred_term: endothelial cell
    term:
      id: CL:0000115
      label: endothelial cell
  evidence:
  - reference: PMID:21885404
    reference_title: Identification of a disease-defining gene fusion in epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This WWTR1/CAMTA1 gene fusion is under the transcriptional control of the
      WWTR1 promoter and encodes a putative chimeric transcription factor that
      joins the amino terminus of WWTR1, a protein that is highly expressed in
      endothelial cells, in-frame to the carboxyl terminus of CAMTA1, a protein
      that is normally expressed only in brain.
    explanation: >-
      The discovery paper establishes the fusion architecture and the
      promoter-switch mechanism that drives inappropriate CAMTA1 expression in
      endothelium.
  - reference: PMID:21584898
    reference_title: A novel WWTR1-CAMTA1 gene fusion is a consistent abnormality in epithelioid hemangioendothelioma of different anatomic sites.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Our results demonstrate the presence of a WWTR1-CAMTA1 fusion in all EHE
      tested from bone, soft tissue, and visceral location (liver, lung) in
      keeping with a unique and specific pathological entity.
    explanation: >-
      Independent confirmation that the fusion is consistently present across
      anatomic sites and grades, establishing it as the disease-defining lesion.
  downstream:
  - target: Constitutive Nuclear TAZ-TEAD Transcriptional Activation
    description: >-
      The fusion protein escapes Hippo-pathway cytoplasmic restraint and acts as
      a constitutively active TAZ.

- name: YAP1-TFE3 Fusion Oncogene Formation
  biological_scale: MOLECULAR
  role: trigger
  subtypes:
  - YAP1-TFE3
  description: >-
    In a minority of tumours a YAP1-TFE3 fusion replaces WWTR1-CAMTA1. YAP1
    exon 1 is fused to TFE3 exon 4, retaining the TFE3 bHLH-leucine-zipper
    DNA-binding and transactivation domains while discarding the YAP1 S127
    14-3-3 site, WW domain and C-terminal transactivation domain. YAP1
    therefore supplies a strong endothelial promoter that misexpresses an
    otherwise intact MiT-family transcription factor. These tumours are
    WWTR1-CAMTA1-negative, CAMTA1-immunonegative and strongly TFE3-positive.
  genes:
  - preferred_term: YAP1
    term:
      id: hgnc:16262
      label: YAP1
  - preferred_term: TFE3
    term:
      id: hgnc:11752
      label: TFE3
  gene_products:
  - preferred_term: YAP1-TFE3 fusion oncoprotein
    term:
      id: NCIT:C177676
      label: YAP1/TFE3 Fusion Protein
  molecular_functions:
  - preferred_term: DNA-binding transcription factor activity
    modifier: ABNORMAL
    term:
      id: GO:0003700
      label: DNA-binding transcription factor activity
  evidence:
  - reference: PMID:23737213
    reference_title: Novel YAP1-TFE3 fusion defines a distinct subset of epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In summary, we are reporting a novel subset of EHE occurring in young
      adults, showing a distinct phenotype and YAP1-TFE3 fusions.
    explanation: >-
      Defines the alternative fusion driver and its distinct clinicopathologic
      subset.
  - reference: PMID:34381186
    reference_title: "YAP1-TFE3-fused hemangioendothelioma: a multi-institutional clinicopathologic study of 24 genetically-confirmed cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All cases were confirmed by molecular testing to harbor YAP1-TFE3 gene
      fusions: majority with YAP1 exon 1 fused to TFE3 exon 4 (88%), or less
      commonly, TFE3 exon 6 (12%).
    explanation: >-
      Establishes the recurrent breakpoint structure of the YAP1-TFE3 fusion in
      a genetically confirmed multi-institutional series.
  downstream:
  - target: Epithelioid Endothelial Neoplastic Proliferation
    description: >-
      Misexpressed TFE3 drives an alternative oncogenic transcriptional program
      producing a morphologically distinct epithelioid vascular tumour.

- name: Constitutive Nuclear TAZ-TEAD Transcriptional Activation
  biological_scale: MOLECULAR
  role: central_effector
  description: >-
    Normally, Hippo-pathway LATS kinases phosphorylate TAZ, promoting 14-3-3
    binding and cytoplasmic sequestration. TAZ-CAMTA1 loses the C-terminal
    regulatory elements and gains CAMTA1-mediated nuclear localisation, so it
    is largely relieved of that restraint and behaves as a constitutively
    activated TAZ. Importantly, the restraint is attenuated rather than
    abolished: LATS1/2 deletion and mutation of the LATS phosphosites still
    increase the fusion's nuclear localisation and target-gene output, so
    TAZ-CAMTA1 remains susceptible to Hippo regulation - which is why
    interventions that drive YAP/TAZ phosphorylation and cytoplasmic
    sequestration (e.g. statins) can inhibit it. The nuclear fusion drives
    TEAD-dependent transcription; disruption of the fusion-TEAD interaction or
    expression of a dominant-negative TEAD blocks transformation, identifying
    this node as the central effector of the disease and the primary
    fusion-directed drug target.
  molecular_functions:
  - preferred_term: transcription coactivator activity
    modifier: INCREASED
    term:
      id: GO:0003713
      label: transcription coactivator activity
  biological_processes:
  - preferred_term: hippo signaling
    modifier: DECREASED
    term:
      id: GO:0035329
      label: hippo signaling
  - preferred_term: positive regulation of transcription by RNA polymerase II
    modifier: INCREASED
    term:
      id: GO:0045944
      label: positive regulation of transcription by RNA polymerase II
  evidence:
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This fusion gene has been observed in almost all reported EHE cases and
      functions as a constitutively activated TAZ.
    explanation: >-
      States directly that the fusion behaves as a constitutively active TAZ,
      the defining molecular consequence modelled by this node.
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      disruption of the TAZ-CAMTA1-TEAD interaction or ectopic expression of a
      dominant negative TEAD in vivo inhibits TAZ-CAMTA1-mediated transformation
    explanation: >-
      Functional in vivo demonstration that TEAD engagement is required for
      fusion-driven transformation, establishing TAZ-TEAD activity as the
      central effector.
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      We conclude that, despite its gain-of-function activity, TAZ-CAMTA1 is
      still susceptible to regulation by Hippo signaling.
    explanation: >-
      Qualifies the "constitutive" framing: the fusion escapes most but not all
      Hippo restraint, which is the basis for pharmacological strategies that
      restore cytoplasmic sequestration. Support is PARTIAL because it tempers
      rather than simply confirming the node's central claim.
  downstream:
  - target: Endothelial Cell Reprogramming
    description: >-
      Sustained TEAD-dependent transcription rewires the endothelial gene
      expression program.
  - target: CTGF-Integrin-RAS-MAPK Signaling
    description: >-
      CTGF is a tumorigenic transcriptional target of the fusion, relaying its
      output into RAS-MAPK signalling.
  - target: mTORC1 Pathway Activation
    description: >-
      YAP/TAZ activity converges on mTOR complex 1, the biologic rationale for
      mTOR inhibition in EHE.

- name: mTORC1 Pathway Activation
  biological_scale: MOLECULAR
  role: effector
  description: >-
    YAP/TAZ activity - and hence the EHE fusion oncoproteins - converges on
    activation of mTOR complex 1. This is the biologic rationale for sirolimus,
    the systemic agent with the strongest disease-specific activity signal in
    EHE. Sirolimus acts on this node, inhibiting phosphorylation of the mTORC1
    substrate 4E-BP1 and, downstream of that, suppressing the ATF4/ATF5-driven
    release of GDF-15 that tracks disease aggressiveness.
  biological_processes:
  - preferred_term: TORC1 signaling
    modifier: INCREASED
    term:
      id: GO:0038202
      label: TORC1 signaling
  evidence:
  - reference: PMID:33107994
    reference_title: Sirolimus for patients with progressive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      In vitro studies have demonstrated that YAP/TAZ, the oncogenic fusions of
      which are the molecular hallmark of EHE, are involved in activation of
      mTOR complex 1.
    explanation: >-
      States the mechanistic link from the EHE fusion effectors to mTORC1.
      Evidence source is OTHER because this is an editorial commentary
      summarising the in vitro literature.
  - reference: PMID:39283723
    reference_title: GDF-15 Predicts Epithelioid Hemangioendothelioma Aggressiveness and Is Downregulated by Sirolimus through ATF4/ATF5 Suppression.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      This reduction was attributed to the drug-induced inhibition of
      phosphorylation/activation of 4E-BP1 and subsequent downregulation of the
      GDF-15 transcription factors ATF4 and ATF5.
    explanation: >-
      Demonstrates that sirolimus engages this node in EHE models, with 4E-BP1
      phosphorylation as the proximal mTORC1 readout.
  downstream:
  - target: Epithelioid Endothelial Neoplastic Proliferation
    description: >-
      mTORC1-driven cap-dependent translation supports tumour cell growth and
      proliferation.

- name: Endothelial Cell Reprogramming
  biological_scale: CELLULAR
  role: effector
  description: >-
    Expression of TAZ-CAMTA1 in endothelial cells initiates an angiogenic and
    regenerative-like transcriptional program and is by itself sufficient to
    generate vascular tumours with the distinctive features of EHE. Two
    independent conditional mouse models - endogenous-locus knock-in and
    endothelial-specific expression - reproduce human EHE, and inhibiting the
    fusion causes established tumours to regress, establishing continued
    oncogene dependence (oncogene addiction).
  cell_types:
  - preferred_term: blood vessel endothelial cell
    term:
      id: CL:0000071
      label: blood vessel endothelial cell
  biological_processes:
  - preferred_term: angiogenesis
    modifier: INCREASED
    term:
      id: GO:0001525
      label: angiogenesis
  evidence:
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      we show that TAZ-CAMTA1 expression in endothelial cells is sufficient to
      drive the formation of vascular tumors with the distinctive features of
      EHE, and inhibition of TAZ-CAMTA1 results in the regression of these
      vascular tumors
    explanation: >-
      Demonstrates both sufficiency in the endothelial compartment and ongoing
      dependence of established tumours on the fusion.
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      We show that TAZ-CAMTA1 initiates an angiogenic and regenerative-like
      transcriptional program in endothelial cells
    explanation: >-
      Characterises the reprogrammed transcriptional state of the endothelial
      cell of origin.
  - reference: PMID:33766982
    reference_title: WWTR1(TAZ)-CAMTA1 gene fusion is sufficient to dysregulate YAP/TAZ signaling and drive epithelioid hemangioendothelioma tumorigenesis.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      These mice develop EHE tumors that are indistinguishable from human EHE
      clinically, histologically, immunohistochemically, and genetically.
    explanation: >-
      An independent endogenous-locus mouse model confirms that the fusion alone
      recapitulates human EHE with exquisite tissue specificity.
  downstream:
  - target: Epithelioid Endothelial Neoplastic Proliferation
    description: >-
      Reprogrammed endothelial cells form the epithelioid neoplastic population
      of EHE.

- name: CTGF-Integrin-RAS-MAPK Signaling
  biological_scale: MOLECULAR
  role: effector
  conforms_to: "sustaining_proliferative_signaling#Constitutive Mitogenic Pathway Activation"
  description: >-
    Connective tissue growth factor (CTGF/CCN2) is a tumorigenic
    transcriptional target of TAZ-CAMTA1. Secreted CTGF binds integrin
    alphaIIbbeta3, sustaining anchorage-independent proliferation, and drives
    RAS and MAPK signalling; CTGF knockdown reduces pathway activity and MEK
    inhibition with PD 0325901 or trametinib abolishes fusion-driven
    anchorage-independent growth and xenograft expansion. This is the
    disease-specific route by which a Hippo-pathway lesion converges on the
    conserved constitutive-mitogenic-signalling hallmark node, and it supplied
    the preclinical rationale for the phase II trametinib trial. Model caveat:
    this arm was mapped in TAZ-CAMTA1-transformed NIH3T3 fibroblasts and their
    xenografts, not in native human EHE endothelium, so its fidelity to the
    endothelial disease is an open question (see the linked
    HUMAN_MODEL_MISMATCH discussion).
  biological_processes:
  - preferred_term: Ras protein signal transduction
    modifier: INCREASED
    term:
      id: GO:0007265
      label: Ras protein signal transduction
  - preferred_term: ERK1 and ERK2 cascade
    modifier: INCREASED
    term:
      id: GO:0070371
      label: ERK1 and ERK2 cascade
  - preferred_term: anoikis
    modifier: DECREASED
    term:
      id: GO:0043276
      label: anoikis
  evidence:
  - reference: PMID:35443056
    reference_title: The TAZ-CAMTA1 Fusion Protein Promotes Tumorigenesis via Connective Tissue Growth Factor and Ras-MAPK Signaling in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      We identified connective tissue growth factor (CTGF) as a tumorigenic
      transcriptional target of TC.
    explanation: >-
      Identifies CTGF as the direct tumorigenic transcriptional target of the
      fusion oncoprotein.
  - reference: PMID:35443056
    reference_title: The TAZ-CAMTA1 Fusion Protein Promotes Tumorigenesis via Connective Tissue Growth Factor and Ras-MAPK Signaling in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Collectively, our results demonstrated that CTGF and the Ras-MAPK
      signaling cascade are essential for TC-mediated tumorigenesis.
    explanation: >-
      Establishes RAS-MAPK dependence, the conformance basis for the
      sustaining-proliferative-signaling hallmark node.
  downstream:
  - target: Epithelioid Endothelial Neoplastic Proliferation
    description: >-
      RAS-MAPK output sustains anchorage-independent tumour cell proliferation.

- name: Epithelioid Endothelial Neoplastic Proliferation
  biological_scale: TISSUE
  role: consequence
  description: >-
    The reprogrammed endothelial cells form the characteristic lesion: cords,
    strands and nests of epithelioid cells embedded in a myxohyaline
    (myxochondroid to sclerotic) stroma, with primitive intracytoplasmic
    vascular lumina ("blister cells") rather than well-formed vascular
    channels. Tumours retain endothelial differentiation (CD31/PECAM1 and ERG
    positive) and grow in an infiltrative pattern in liver, lung, bone and soft
    tissue.
  cell_types:
  - preferred_term: neoplastic epithelioid endothelial cell
    term:
      id: CL:0000115
      label: endothelial cell
  biological_processes:
  - preferred_term: cell population proliferation
    modifier: INCREASED
    term:
      id: GO:0008283
      label: cell population proliferation
  locations:
  - preferred_term: liver
    term:
      id: UBERON:0002107
      label: liver
  - preferred_term: lung
    term:
      id: UBERON:0002048
      label: lung
  - preferred_term: bone tissue
    term:
      id: UBERON:0002481
      label: bone tissue
  evidence:
  - reference: PMID:32913529
    reference_title: Fluorescence in situ hybridization for WWTR1-CAMTA1 has higher sensitivity and specificity for epithelioid hemangioendothelioma diagnosis.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Histologically, EHE typically showed a mucous hyaline or cartilaginous
      stroma, often forming a primitive vascular lumen, and expressed vascular
      endothelial markers.
    explanation: >-
      Describes the defining tissue-level architecture and retained endothelial
      phenotype of the neoplastic proliferation.
  downstream:
  - target: Multifocal Dissemination and Serosal Involvement
    description: >-
      The tumour is frequently multifocal at diagnosis and disseminates to
      lymph nodes, other viscera and serosal surfaces.

- name: CDKN2A Loss and Tumour Progression
  biological_scale: MOLECULAR
  role: modifier
  conforms_to: "evading_growth_suppressors#Tumor Suppressor Inactivation"
  description: >-
    Roughly half of EHEs carry a secondary genetic alteration beyond the
    defining fusion; loss of the CDKN2A tumour suppressor is the most common.
    In a conditional mouse model, adding Cdkn2a knockout to the TAZ-CAMTA1
    allele produced earlier tumour-related morbidity and mortality and enhanced
    tumour cell proliferation, identifying CDKN2A loss as a progression event
    rather than an initiating one. This is the disease-specific substitution
    for the conserved tumour-suppressor-inactivation trigger of the
    evading-growth-suppressors hallmark.
  genes:
  - preferred_term: CDKN2A
    term:
      id: hgnc:1787
      label: CDKN2A
  biological_processes:
  - preferred_term: cell cycle checkpoint signaling
    modifier: DECREASED
    term:
      id: GO:0000075
      label: cell cycle checkpoint signaling
  - preferred_term: negative regulation of cell cycle
    modifier: DECREASED
    term:
      id: GO:0045786
      label: negative regulation of cell cycle
  evidence:
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Sequencing of human tumors has, however, identified additional secondary
      mutations in approximately 50% of EHE, most commonly the loss of tumor
      suppressor CDKN2A.
    explanation: >-
      Establishes CDKN2A loss as the most frequent secondary alteration in human
      EHE. Note that this statement summarises prior sequencing studies in the
      paper's Purpose section rather than reporting this paper's own cohort.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Memorial Sloan Kettering-IMPACT targeted DNA sequencing was performed to
      identify secondary genetic alterations showing more than half of tumors
      had a genetic alteration beyond the disease-defining gene fusion.
    explanation: >-
      Primary sequencing cohort corroborating that secondary alterations occur
      in more than half of EHE tumours.
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Loss of Cdkn2a within EHE was associated with more aggressive disease, as
      displayed by earlier tumor-related morbidity/mortality and enhanced tumor
      cell proliferation.
    explanation: >-
      Functional mouse evidence that CDKN2A loss accelerates EHE progression,
      supporting its role as a modifier node and its route through enhanced
      proliferation.
  downstream:
  - target: Epithelioid Endothelial Neoplastic Proliferation
    description: >-
      Loss of the CDKN2A-enforced cell-cycle checkpoint removes an
      antiproliferative brake, enhancing tumour cell proliferation. This is the
      disease-specific instance of the module edge Tumor Suppressor
      Inactivation to Loss of Cell-Cycle Checkpoint Control.

- name: Multifocal Dissemination and Serosal Involvement
  biological_scale: ORGANISM
  role: consequence
  conforms_to: "invasion_and_metastasis#Metastatic Colonization"
  description: >-
    EHE is characteristically multifocal, and multifocality is best interpreted
    as early metastatic colonisation rather than independent primary tumours.
    Successful colonisation of lung, liver, bone, lymph nodes and - most
    ominously - pleural and peritoneal surfaces determines outcome. Pleural
    involvement and lymph node metastasis confer an aggressive course akin to
    high-grade sarcoma, and serosal effusion is the single most consistent
    adverse prognostic feature.
  biological_processes:
  - preferred_term: cell population proliferation
    modifier: INCREASED
    term:
      id: GO:0008283
      label: cell population proliferation
  - preferred_term: cell migration
    modifier: INCREASED
    term:
      id: GO:0016477
      label: cell migration
  locations:
  - preferred_term: pleura
    term:
      id: UBERON:0000977
      label: pleura
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Multifocality, pleural involvement, lymph node or distant metastases had a
      significantly worse outcome.
    explanation: >-
      Establishes multifocal and metastatic colonisation, especially pleural, as
      the outcome-determining step in EHE.
  - reference: PMID:37444414
    reference_title: "Prognostic Factors in Epithelioid Hemangioendothelioma: Analysis of a Nationwide Molecularly/Immunohistochemically Confirmed Cohort of 57 Cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      No survival differences were observed between multifocal and metastatic
      disease, suggesting that multifocality represents early metastases
    explanation: >-
      Clinical support for interpreting EHE multifocality as metastatic
      colonisation.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      multifocal liver epithelioid hemangioendothelioma represents a clonal
      process, demonstrating that separate lesions harbor fusions with an
      identical genomic breakpoint, in keeping with a single clone, i.e.
      metastatic dissemination confined to one organ
    explanation: >-
      Direct molecular (shared-breakpoint) evidence that multifocal hepatic EHE
      is clonal metastatic dissemination rather than independent multicentric
      transformation. This is the strongest basis for conformance to
      invasion_and_metastasis#Metastatic Colonization.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Among the entire WWTR1-CAMTA1 positive cohort, 29 patients (35%) had
      pleural involvement, which most commonly (83% of pleural cases) was
      present at diagnosis.
    explanation: >-
      Quantifies pleural (serosal) colonisation in a molecularly confirmed
      cohort and shows it is usually present from the outset.

histopathology:
- name: Epithelioid Cells in Myxohyaline Stroma
  finding_term:
    preferred_term: myxohyaline (mucoid) stroma
    term:
      id: NCIT:C35999
      label: Mucoid Stroma Formation
  diagnostic: true
  subtype: WWTR1-CAMTA1
  description: >-
    Conventional EHE shows cords, strands and single files of epithelioid
    endothelial cells set in a distinctive myxoid, myxochondroid or sclerotic
    (hyaline) extracellular stroma, generally without well-formed vascular
    channels.
  evidence:
  - reference: PMID:32913529
    reference_title: Fluorescence in situ hybridization for WWTR1-CAMTA1 has higher sensitivity and specificity for epithelioid hemangioendothelioma diagnosis.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Histologically, EHE typically showed a mucous hyaline or cartilaginous
      stroma, often forming a primitive vascular lumen, and expressed vascular
      endothelial markers.
    explanation: >-
      Directly documents the myxohyaline/cartilaginous stroma as the typical
      histologic background of EHE.
- name: Intracytoplasmic Vascular Lumina (Blister Cells)
  finding_term:
    preferred_term: intracytoplasmic vacuolation forming primitive vascular lumina
    term:
      id: NCIT:C96302
      label: Cytoplasmic Vacuolation
  diagnostic: true
  description: >-
    Rather than forming mature vascular channels, conventional EHE cells
    contain immature intracytoplasmic lumina or vacuoles ("blister cells"),
    which represent primitive vascular differentiation and are a key
    discriminator from epithelioid haemangioma and epithelioid angiosarcoma.
  evidence:
  - reference: PMID:21584898
    reference_title: A novel WWTR1-CAMTA1 gene fusion is a consistent abnormality in epithelioid hemangioendothelioma of different anatomic sites.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      the most important histologic criteria applied to distinguish EHE from
      both EH and E-AS was the lack of well-formed vascular channel formation in
      EHE, with only immature, intra-cytoplasmic lumina being observed
    explanation: >-
      Documents intracytoplasmic lumina as the defining vasoformative feature
      distinguishing EHE from its principal mimics.
- name: Atypical Histology (Size and Mitotic Activity)
  finding_term:
    preferred_term: increased mitotic activity
    term:
      id: NCIT:C163732
      label: Increased Mitotic Activity
  description: >-
    A subset of EHE shows atypical histology, defined as at least two of high
    mitotic activity, high nuclear grade and coagulative necrosis. Together
    with tumour size greater than 30 mm this defines a validated three-tier
    risk stratification.
  evidence:
  - reference: PMID:33729740
    reference_title: "Clinicopathologic Characterization of Epithelioid Hemangioendothelioma in a Series of 62 Cases: A Proposal of Risk Stratification and Identification of a Synaptophysin-positive Aggressive Subset."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Twenty-two tumors (35.5%) demonstrated atypical histology, defined by
      having at least 2 of the following 3 findings: high mitotic activity
      (>1/2 mm2), high nuclear grade, and coagulative necrosis.
    explanation: >-
      Defines the atypical-histology criteria and their frequency in a
      molecularly confirmed cohort.
- name: Vasoformative Architecture with Eosinophilic Cytoplasm
  finding_term:
    preferred_term: nested and pseudoalveolar growth pattern
    term:
      id: NCIT:C35892
      label: Nested Pattern
  subtype: YAP1-TFE3
  description: >-
    YAP1-TFE3-fused tumours differ morphologically from conventional EHE:
    solid sheets of coalescing nests, a pseudoalveolar/(pseudo)vasoformative
    pattern, and discohesive strands in myxohyaline stroma, with large
    epithelioid cells having abundant glassy eosinophilic to vacuolated
    cytoplasm.
  evidence:
  - reference: PMID:34381186
    reference_title: "YAP1-TFE3-fused hemangioendothelioma: a multi-institutional clinicopathologic study of 24 genetically-confirmed cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      we describe three dominant architectural patterns: solid sheets of
      coalescing nests, pseudoalveolar and (pseudo)vasoformative pattern, and
      discohesive strands and clusters of cells set in a myxoid to myxohyaline
      stroma.
    explanation: >-
      Characterises the distinct architecture of the YAP1-TFE3 subtype.

phenotypes:
- category: Neoplastic
  name: Vascular Sarcoma
  diagnostic: true
  description: >-
    EHE is a malignant vascular (endothelial) neoplasm whose clinical behaviour
    is intermediate between benign haemangioma and high-grade angiosarcoma.
    Note that "intermediate" here describes behaviour: the current WHO
    soft-tissue classification places EHE in the malignant vascular tumour
    category, not the intermediate/rarely-metastasising one.
  phenotype_term:
    preferred_term: Sarcoma
    term:
      id: HP:0100242
      label: Sarcoma
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Epithelioid hemangioendothelioma is a low-grade malignant vascular tumor
      with an intermediate clinical behavior between benign hemangiomas and
      high-grade angiosarcomas.
    explanation: >-
      Defines the core disease identity as a low-grade malignant vascular
      neoplasm.
- category: Hepatobiliary
  name: Hepatic Involvement
  description: >-
    The liver is one of the most frequent sites of EHE, commonly presenting as
    multifocal peripheral nodules; hepatic disease carries an intermediate
    prognosis.
  phenotype_term:
    preferred_term: Neoplasm of the liver
    term:
      id: HP:0002896
      label: Neoplasm of the liver
  evidence:
  - reference: PMID:37260142
    reference_title: "Incidence, demographics, and survival of malignant hemangioendothelioma in the United States."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Skin and connective tissues were the most common presenting sites (33.4%),
      followed by liver (24.5%), lung (17.6%), and bone (12.5%).
    explanation: >-
      Establishes the liver as a major site of disease. Support is PARTIAL and
      no `frequency` band is asserted because this figure is the distribution of
      *presenting primary site* in the broader malignant-hemangioendothelioma
      registry group, not the fraction of EHE patients with hepatic
      involvement - which is higher, since most patients are multifocal.
- category: Respiratory
  name: Pulmonary Involvement
  description: >-
    Pulmonary EHE - historically called intravascular bronchioloalveolar tumour
    - typically presents as multiple bilateral nodules and/or pleural
    thickening, and lung involvement carries the least favourable
    site-specific survival.
  phenotype_term:
    preferred_term: Neoplasm of the lung
    term:
      id: HP:0100526
      label: Neoplasm of the lung
  frequency: FREQUENT
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In combination with other metastatic foci, such as the liver or bone, 49
      cases (59%) involved the lung at diagnosis.
    explanation: >-
      In a cohort of 83 molecularly confirmed WWTR1-CAMTA1-positive EHE
      patients, 59% had lung involvement at diagnosis, mapping to the FREQUENT
      (30-79%) band. This directly measures organ involvement in EHE, unlike
      registry primary-site distributions.
- category: Musculoskeletal
  name: Skeletal Involvement
  description: >-
    Bone EHE may present with pain and lytic lesions and can be multifocal
    within a single bone or across the skeleton.
  phenotype_term:
    preferred_term: Neoplasm of the skeletal system
    term:
      id: HP:0010622
      label: Neoplasm of the skeletal system
  evidence:
  - reference: PMID:37260142
    reference_title: "Incidence, demographics, and survival of malignant hemangioendothelioma in the United States."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Skin and connective tissues were the most common presenting sites (33.4%),
      followed by liver (24.5%), lung (17.6%), and bone (12.5%).
    explanation: >-
      Establishes bone as a recognised site of disease. Support is PARTIAL and
      no `frequency` band is asserted because this figure is the distribution of
      presenting primary site in the broader malignant-hemangioendothelioma
      registry group, not the fraction of EHE patients with skeletal
      involvement.
- category: Musculoskeletal
  name: Soft Tissue Involvement
  description: >-
    Soft tissue (with skin) is the single most common site of primary EHE and,
    when unifocal, the presentation most amenable to curative resection and
    carrying the most favourable outcome.
  phenotype_term:
    preferred_term: Soft tissue neoplasm
    term:
      id: HP:0031459
      label: Soft tissue neoplasm
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A total of 17 patients (20% of the whole WWTR1-CAMTA1-positive cohort) had
      primary soft tissue tumors, all but one had solitary disease.
    explanation: >-
      Quantifies primary soft tissue disease at 20% of a molecularly confirmed
      EHE cohort, mapping to the OCCASIONAL (5-29%) band, and notes its
      characteristically solitary presentation.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Soft tissue epithelioid hemangioendothelioma were frequently solitary,
      followed an uneventful clinical course being often managed with curative
      surgery.
    explanation: >-
      Supports the favourable, surgically curable behaviour of soft tissue EHE.
- category: Respiratory
  name: Pleural Effusion
  description: >-
    Pleural effusion reflects serosal spread and is the most consistently
    reported adverse prognostic feature in EHE; patients with serosal effusion
    have markedly shorter progression-free and overall survival and respond
    poorly to sirolimus.
  phenotype_term:
    preferred_term: Pleural effusion
    term:
      id: HP:0002202
      label: Pleural effusion
  evidence:
  - reference: PMID:33107985
    reference_title: "Activity of sirolimus in patients with progressive epithelioid hemangioendothelioma: A case-series analysis within the Italian Rare Cancer Network."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Serosal effusions are confirmed as an unfavorable prognostic sign
      associated with short survival, and sirolimus displays limited activity in
      this subgroup.
    explanation: >-
      Supports serosal effusion as both a clinical manifestation and a
      treatment-resistant adverse prognostic feature.
- category: Gastrointestinal
  name: Ascites
  description: >-
    Peritoneal involvement may produce ascites, the abdominal counterpart of
    pleural effusion and a similarly adverse serosal manifestation.
  phenotype_term:
    preferred_term: Ascites
    term:
      id: HP:0001541
      label: Ascites
  evidence:
  - reference: PMID:33107994
    reference_title: Sirolimus for patients with progressive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      It has been demonstrated that patients with effusions, either pleural or
      peritoneal, have worse outcomes.
    explanation: >-
      Names peritoneal effusion explicitly as an adverse EHE feature. Evidence
      source is OTHER because this is an editorial commentary summarising the
      cohort literature.
  - reference: PMID:33107985
    reference_title: "Activity of sirolimus in patients with progressive epithelioid hemangioendothelioma: A case-series analysis within the Italian Rare Cancer Network."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      at baseline, 13 of 38 patients had the presence of serosal effusions and
      systemic symptoms
    explanation: >-
      Documents serosal effusions in roughly a third of patients with advanced
      EHE. Support is PARTIAL because the report aggregates pleural and
      peritoneal effusion rather than reporting ascites separately.
- category: Constitutional
  name: Systemic Symptoms
  description: >-
    Constitutional deterioration - weight loss, fatigue and systemic symptoms -
    accompanies aggressive disease and, with anaemia and new serosal
    involvement, marks the transition to a poor-prognosis phenotype.
  phenotype_term:
    preferred_term: Weight loss
    term:
      id: HP:0001824
      label: Weight loss
  evidence:
  - reference: PMID:33107994
    reference_title: Sirolimus for patients with progressive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      pulmonary disease, weight loss, anemia, and pulmonary hemorrhage were
      correlated with poorer patient outcomes
    explanation: >-
      Names weight loss explicitly as an EHE feature correlated with poorer
      outcome. Evidence source is OTHER because this is an editorial commentary
      summarising the cohort literature.
  - reference: PMID:33107985
    reference_title: "Activity of sirolimus in patients with progressive epithelioid hemangioendothelioma: A case-series analysis within the Italian Rare Cancer Network."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      at baseline, 13 of 38 patients had the presence of serosal effusions and
      systemic symptoms
    explanation: >-
      Documents systemic symptoms as a recognised baseline feature of advanced
      EHE. Support is PARTIAL because the source does not itemise weight loss
      separately from other constitutional symptoms.
- category: Hematologic
  name: Anemia
  description: >-
    Anaemia accompanies aggressive EHE and, together with new serosal
    involvement and constitutional decline, is one of the practical triggers
    for abandoning observation in favour of systemic therapy.
  phenotype_term:
    preferred_term: Anemia
    term:
      id: HP:0001903
      label: Anemia
  evidence:
  - reference: PMID:33107994
    reference_title: Sirolimus for patients with progressive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      pulmonary disease, weight loss, anemia, and pulmonary hemorrhage were
      correlated with poorer patient outcomes
    explanation: >-
      Names anaemia as an EHE feature correlated with poorer outcome. Evidence
      source is OTHER because this is an editorial commentary summarising the
      cohort literature.
- category: Systemic
  name: Metastatic Disease
  description: >-
    Distant metastasis is the rule rather than the exception in
    fusion-confirmed EHE, and survival falls sharply with extent of disease at
    presentation.
  phenotype_term:
    preferred_term: metastatic disease
  notes: >-
    No `term:` binding is asserted. HPO has no generic metastasis term
    (HP:0200059 is Metastatic angiosarcoma, a different disease) and the
    available generic HP:0002664 Neoplasm would encode nothing beyond "this is
    a cancer"; metastasis is better represented from NCIT/SNOMED, which is not
    in scope for `phenotype_term`.
  frequency: FREQUENT
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The majority of patients (n=61; 73%) had multifocal primary disease at the
      time of diagnosis and 70% (n=58) developed distant metastases.
    explanation: >-
      In a molecularly confirmed EHE cohort, 70% developed distant metastases
      and 73% were multifocal at diagnosis, mapping to the FREQUENT (30-79%)
      band.
  - reference: PMID:37260142
    reference_title: "Incidence, demographics, and survival of malignant hemangioendothelioma in the United States."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      At presentation, 36.4% of patients had localized disease; 21.6% presented
      with regional and 41.7% with distant metastases.
    explanation: >-
      Population-registry corroboration that a plurality of patients already
      have distant spread at diagnosis.

biochemical:
- name: Nuclear CAMTA1 Immunohistochemistry
  biomarker_term:
    preferred_term: WWTR1-CAMTA1 fusion protein expression
    term:
      id: NCIT:C121670
      label: WWTR1-CAMTA1 Fusion Protein Expression
  notes: >-
    Nuclear CAMTA1 staining is a widely used surrogate for the WWTR1-CAMTA1
    fusion, reported positive in about 86% of EHE and essentially absent in
    other epithelioid mesenchymal neoplasms. Specificity is not absolute - an
    independent series found nuclear CAMTA1 positivity in 6 of 37 histologic
    mimics - so FISH or RNA-based fusion detection remains the more reliable
    confirmatory test.
  evidence:
  - reference: PMID:26414223
    reference_title: Nuclear Expression of CAMTA1 Distinguishes Epithelioid Hemangioendothelioma From Histologic Mimics.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In total, 51/59 cases (86%) of EHE showed diffuse nuclear staining for
      CAMTA1, including 44/48 cases (92%) with conventional histology and 7/11
      cases (64%) with "malignant" histology.
    explanation: >-
      Quantifies the sensitivity of nuclear CAMTA1 immunohistochemistry for EHE.
  - reference: PMID:32913529
    reference_title: Fluorescence in situ hybridization for WWTR1-CAMTA1 has higher sensitivity and specificity for epithelioid hemangioendothelioma diagnosis.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Twelve of the 15 EHE samples showed positive nuclear CAMTA1 expression
      with immunohistochemistry, whereas six of the 37 histologic mimics showed
      positive nuclear expression.
    explanation: >-
      Tempers the claim: CAMTA1 immunohistochemistry is useful but imperfectly
      specific, which is why molecular fusion confirmation is preferred.
- name: WWTR1-CAMTA1 Fusion Detection by FISH or RNA Sequencing
  biomarker_term:
    preferred_term: WWTR1/CAMTA1 fusion gene
    term:
      id: NCIT:C121672
      label: WWTR1/CAMTA1 Fusion Gene
  notes: >-
    Break-apart FISH or RNA-based fusion assays are the diagnostic gold
    standard. A negative CAMTA1 immunostain in morphologically convincing
    disease should prompt an RNA-based fusion assay rather than exclusion of
    EHE, since the tumour may carry the YAP1-TFE3 fusion or a variant WWTR1
    partner.
  evidence:
  - reference: PMID:32913529
    reference_title: Fluorescence in situ hybridization for WWTR1-CAMTA1 has higher sensitivity and specificity for epithelioid hemangioendothelioma diagnosis.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      FISH detected a red-green signal fusion in 14 of the 15 cases of EHE, but
      in none of the 10 vascular tumors.
    explanation: >-
      Quantifies FISH sensitivity and specificity for the defining fusion.
- name: Endothelial Marker Immunohistochemistry (CD31, ERG, CD34)
  notes: >-
    Retained endothelial differentiation is confirmed with CD31/PECAM1 and ERG,
    with CD34 variably positive; focal cytokeratin expression may occur and
    must not be misread as carcinoma.
  evidence:
  - reference: PMID:34381186
    reference_title: "YAP1-TFE3-fused hemangioendothelioma: a multi-institutional clinicopathologic study of 24 genetically-confirmed cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Immunohistochemically, ERG, CD31, and TFE3 were consistently expressed,
      while expression of CD34 (83%) and cytokeratin AE1/AE3 (20%) was variable.
    explanation: >-
      Documents the endothelial immunophenotype and the variable cytokeratin
      expression that can cause diagnostic confusion.
- name: Circulating GDF-15
  biomarker_term:
    preferred_term: growth/differentiation factor 15
    term:
      id: NCIT:C34026
      label: Growth/Differentiation Factor 15
  notes: >-
    An emerging, not yet validated, circulating biomarker. GDF-15 is released
    by EHE cells, correlates with disease aggressiveness across two independent
    patient series, and is downregulated by sirolimus via inhibition of 4E-BP1
    phosphorylation and suppression of the transcription factors ATF4 and ATF5.
    It is a research biomarker only and is not regulatory-qualified.
  evidence:
  - reference: PMID:39283723
    reference_title: GDF-15 Predicts Epithelioid Hemangioendothelioma Aggressiveness and Is Downregulated by Sirolimus through ATF4/ATF5 Suppression.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Blood sample analyses from two independent patient series showed a
      significant correlation between GDF-15 and EHE aggressiveness.
    explanation: >-
      Human evidence linking circulating GDF-15 to EHE aggressiveness.
  - reference: PMID:39283723
    reference_title: GDF-15 Predicts Epithelioid Hemangioendothelioma Aggressiveness and Is Downregulated by Sirolimus through ATF4/ATF5 Suppression.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      This reduction was attributed to the drug-induced inhibition of
      phosphorylation/activation of 4E-BP1 and subsequent downregulation of the
      GDF-15 transcription factors ATF4 and ATF5.
    explanation: >-
      Mechanistic basis for GDF-15 as a pharmacodynamic readout of mTOR
      inhibition in EHE.

genetic:
- name: WWTR1-CAMTA1
  association: Somatic t(1;3)(p36.23;q25.1) gene fusion
  frequency: approximately 90% of cases
  notes: >-
    The disease-defining somatic structural variant. It is not a germline
    Mendelian allele; population allele frequencies are not applicable, and
    there is no established inherited predisposition to EHE.
  gene_term:
    preferred_term: WWTR1
    term:
      id: hgnc:24042
      label: WWTR1
  evidence:
  - reference: PMID:21885404
    reference_title: Identification of a disease-defining gene fusion in epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The gene fusion is present in virtually all EHEs tested but is absent from
      all other vascular neoplasms, demonstrating it to be a disease-defining
      genetic alteration.
    explanation: >-
      Establishes the fusion as necessary and specific for EHE among vascular
      neoplasms.
  - reference: PMID:35740643
    reference_title: "Unraveling the Biology of Epithelioid Hemangioendothelioma, a TAZ-CAMTA1 Fusion Driven Sarcoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Herein, we present a unique TAZ-driven cancer, epithelioid
      hemangioendothelioma (EHE), which harbors a WWTR1(TAZ)-CAMTA1 gene fusion
      in at least 90% of cases.
    explanation: >-
      Review-level source for the approximately 90% prevalence of the fusion.
      Evidence source is OTHER because this is a review article.
- name: CAMTA1
  association: Somatic gene fusion partner (3-prime partner of WWTR1-CAMTA1)
  notes: >-
    CAMTA1 supplies the TIG DNA-binding domain, ankyrin repeats and IQ motifs
    to the fusion, plus a nuclear-localisation function that helps the chimera
    evade Hippo-mediated cytoplasmic restraint. CAMTA1 is normally expressed
    only in brain.
  gene_term:
    preferred_term: CAMTA1
    term:
      id: hgnc:18806
      label: CAMTA1
  evidence:
  - reference: PMID:21885404
    reference_title: Identification of a disease-defining gene fusion in epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Thus, CAMTA1 expression is activated inappropriately through a
      promoter-switch mechanism.
    explanation: >-
      Establishes the mechanism by which the CAMTA1 partner is misexpressed in
      endothelium.
- name: YAP1
  association: Somatic gene fusion (YAP1-TFE3) in a minority subset
  frequency: about 5% of cases (reported from <5% up to ~10% depending on series and diagnostic criteria)
  subtype: YAP1-TFE3
  gene_term:
    preferred_term: YAP1
    term:
      id: hgnc:16262
      label: YAP1
  notes: >-
    YAP1 exon 1 supplies a strong endothelial promoter to an otherwise intact
    TFE3 transcription factor. These tumours are WWTR1-CAMTA1-negative. The
    reported share varies with how strictly the entity is defined and with the
    assay used, so the range rather than a single figure is recorded here.
  evidence:
  - reference: PMID:26414223
    reference_title: Nuclear Expression of CAMTA1 Distinguishes Epithelioid Hemangioendothelioma From Histologic Mimics.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      resulting in the formation of a WWTR1-CAMTA1 fusion gene in approximately
      90% of cases; a small subset (<5%) have a YAP1-TFE3 fusion gene.
    explanation: >-
      Supports the lower bound of the range (<5%) alongside the ~90%
      WWTR1-CAMTA1 figure.
  - reference: PMID:32913529
    reference_title: Fluorescence in situ hybridization for WWTR1-CAMTA1 has higher sensitivity and specificity for epithelioid hemangioendothelioma diagnosis.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Approximately 90% and 5% of EHE cases are associated with the
      WWTR1-CAMTA1 and YAP1/TFE3 fusion gene, respectively.
    explanation: >-
      Independent source giving the central estimate of about 5%.
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      YAP1-TFE3 (in ∼10% of cases)
    explanation: >-
      Supports the upper bound of the reported range. Evidence source is OTHER
      because this is a background statement in the introduction of a
      mechanistic paper rather than its own cohort measurement.
- name: TFE3
  association: Somatic gene fusion partner (3-prime partner of YAP1-TFE3)
  subtype: YAP1-TFE3
  gene_term:
    preferred_term: TFE3
    term:
      id: hgnc:11752
      label: TFE3
  notes: >-
    TFE3 is a MiT-family basic helix-loop-helix leucine-zipper transcription
    factor. As in other TFE3-rearranged tumours (alveolar soft part sarcoma,
    Xp11 translocation renal cell carcinoma), the fusion preserves the
    DNA-binding and transactivation domains, so misexpression alone is
    oncogenic.
  evidence:
  - reference: PMID:23737213
    reference_title: Novel YAP1-TFE3 fusion defines a distinct subset of epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All tumors expressed endothelial markers, as well as strong nuclear TFE3.
    explanation: >-
      Confirms TFE3 protein misexpression as the functional consequence of the
      fusion.
- name: CDKN2A
  association: Secondary somatic loss
  frequency: most common secondary alteration; secondary mutations in about 50% of EHE
  gene_term:
    preferred_term: CDKN2A
    term:
      id: hgnc:1787
      label: CDKN2A
  notes: >-
    A progression rather than initiating event. Roughly 45% of WWTR1-CAMTA1
    tumours have no additional detectable genetic alteration, underscoring the
    sufficiency of the fusion.
  evidence:
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Sequencing of human tumors has, however, identified additional secondary
      mutations in approximately 50% of EHE, most commonly the loss of tumor
      suppressor CDKN2A.
    explanation: >-
      Establishes the frequency and identity of the dominant secondary
      alteration.

prevalence:
- population: United States
  measure_type: ANNUAL_INCIDENCE
  prevalence_class: BELOW_1_IN_1000000
  rate_per_100000: 0.04
  notes: >-
    US Cancer Statistics / SEER combined database, 2001-2017: 0.4 cases per
    million person-years for malignant hemangioendothelioma overall, the
    majority of which are epithelioid. Normalised to 0.04 per 100,000.
  evidence:
  - reference: PMID:37260142
    reference_title: "Incidence, demographics, and survival of malignant hemangioendothelioma in the United States."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The incidence of hemangioendothelioma in the US is 0.4 cases per million
      person-years.
    explanation: >-
      Population-registry incidence estimate for the hemangioendothelioma group
      in which EHE predominates.

progression:
- phase: Heterogeneous course from indolent to rapidly fatal
  notes: >-
    EHE has a markedly heterogeneous natural history. Solitary soft-tissue
    disease is often managed with curative surgery and follows an uneventful
    course, while pleural involvement or lymph node metastasis produces a
    course akin to high-grade sarcoma. Five-year overall survival was 59% in
    the WWTR1-CAMTA1 subset versus 86% in the YAP1-TFE3 subset in a large
    single-institution translocation-positive cohort, and only 22% and 30% in
    patients with pleural disease or lymph node metastases respectively.
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Patients with conventional epithelioid hemangioendothelioma with
      WWTR1-CAMTA1 fusion had a less favorable outcome compared with the
      YAP1-TFE3 subset, the 5-year overall survival being 59% versus 86%,
      respectively.
    explanation: >-
      Quantifies the survival difference between the two molecular subsets.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Patients with pleural disease or lymph node metastases had an aggressive
      clinical course akin to high-grade sarcomas, with 22% and 30%,
      respectively, alive at 5 years, compared with >70% survival rate in
      patients lacking these two adverse factors.
    explanation: >-
      Quantifies the prognostic impact of serosal and nodal involvement.
- phase: Risk stratification by tumour size and histologic atypia
  notes: >-
    A proposed three-tier system combining tumour size greater than 30 mm with
    atypical histology separates low-, intermediate- and high-risk groups with
    5-year overall survival of 100%, 81.8% and 16.9%. It has since been applied
    successfully to unifocal disease in an independent nationwide cohort.
  evidence:
  - reference: PMID:37444414
    reference_title: "Prognostic Factors in Epithelioid Hemangioendothelioma: Analysis of a Nationwide Molecularly/Immunohistochemically Confirmed Cohort of 57 Cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In unifocal tumors, survival could be predicted using the risk
      stratification model of Shibayama et al., dividing the cases into
      low- (n = 4), intermediate- (n = 15) and high- (n = 3) risk groups.
    explanation: >-
      Independent nationwide cohort applying and supporting the proposed
      risk-stratification model in unifocal disease.
  - reference: PMID:33729740
    reference_title: "Clinicopathologic Characterization of Epithelioid Hemangioendothelioma in a Series of 62 Cases: A Proposal of Risk Stratification and Identification of a Synaptophysin-positive Aggressive Subset."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A proposed 3-tiered risk assessment system using these 2 parameters
      significantly stratified the patients into low-risk, intermediate-risk,
      and high-risk groups with 5-year overall survival rates of 100%, 81.8%,
      and 16.9%, respectively (P<0.001).
    explanation: >-
      Provides the validated risk-stratification model and its survival
      separation.
- phase: Stage-dependent survival at presentation
  notes: >-
    Extent of disease at diagnosis strongly predicts outcome, and most deaths
    occur early in the disease course.
  evidence:
  - reference: PMID:37260142
    reference_title: "Incidence, demographics, and survival of malignant hemangioendothelioma in the United States."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Observed survival at 3 years was 79.7%, 70.7%, and 46.0% for patients
      presenting with local, regional, and distant disease and most deaths
      occurred within the first 2 years.
    explanation: >-
      Population-level survival by extent of disease at presentation.

differential_diagnoses:
- name: Epithelioid angiosarcoma
  distinguishing_features:
  - Marked nuclear pleomorphism, brisk mitotic activity and necrosis
  - Negative for both WWTR1-CAMTA1 and YAP1-TFE3 fusions
  - High-grade clinical behaviour rather than intermediate malignancy
  description: >-
    The principal high-grade mimic. No `disease_term` is bound here because
    MONDO currently has no separate class for epithelioid angiosarcoma - it is
    listed as an EXACT synonym of MONDO:0015523 itself (inherited from
    NCIT:C3800). That is an ontology defect: the two are distinct entities with
    different genetics, treatment and prognosis, and binding the EHE term to
    its own differential would propagate the error.
  evidence:
  - reference: PMID:21584898
    reference_title: A novel WWTR1-CAMTA1 gene fusion is a consistent abnormality in epithelioid hemangioendothelioma of different anatomic sites.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This recurrent translocation has not been detected in any of the
      morphologic mimics of EHE, such as EH, E-AS or epithelioid sarcoma-like
      EHE, and thus can serve as a useful molecular diagnostic tool in
      challenging cases.
    explanation: >-
      Demonstrates that the defining fusion is absent from epithelioid
      angiosarcoma, providing the molecular discriminator.
- name: Epithelioid hemangioma
  disease_term:
    preferred_term: epithelioid hemangioma
    term:
      id: MONDO:0021169
      label: epithelioid hemangioma
  distinguishing_features:
  - Benign, lobulated or well-circumscribed lesion
  - Clearly vasoformative with mature open vascular lumina
  - Bland cytomorphology and absence of CAMTA1/WWTR1 rearrangement
  description: >-
    The principal benign mimic. Misdiagnosis of epithelioid haemangioma as EHE
    leads to overtreatment.
  evidence:
  - reference: PMID:21584898
    reference_title: A novel WWTR1-CAMTA1 gene fusion is a consistent abnormality in epithelioid hemangioendothelioma of different anatomic sites.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Particularly important is the lack of CAMTA1 and WWTR1 rearrangements in
      EH, an under-recognized benign tumor, that is often mis-diagnosed as EHE,
      resulting in overtreatment.
    explanation: >-
      Directly documents epithelioid haemangioma as the key benign mimic and the
      clinical cost of confusing it with EHE.
- name: Pseudomyogenic (epithelioid sarcoma-like) hemangioendothelioma
  disease_term:
    preferred_term: pseudomyogenic hemangioendothelioma
    term:
      id: MONDO:0975754
      label: pseudomyogenic hemangioendothelioma
  distinguishing_features:
  - Rhabdomyoblast-like cells with dense eosinophilic cytoplasm
  - Co-expression of endothelial and epithelial markers
  - Typically multiple superficial nodules; negative for the EHE-defining fusions
  evidence:
  - reference: PMID:23737213
    reference_title: Novel YAP1-TFE3 fusion defines a distinct subset of epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A control group, including 18 epithelioid hemangiomas, nine pseudomyogenic
      HE, and three epithelioid angiosarcomas, was also tested.
    explanation: >-
      Confirms pseudomyogenic haemangioendothelioma as a routine comparator that
      was fusion-negative in the TFE3 screen. Support is PARTIAL because the
      abstract does not itemise its alternative driver.
- name: Alveolar soft part sarcoma
  disease_term:
    preferred_term: alveolar soft part sarcoma
    term:
      id: MONDO:0011655
      label: alveolar soft part sarcoma
  distinguishing_features:
  - Also strongly nuclear TFE3-positive with a pseudoalveolar pattern
  - Negative for the endothelial markers CD31 and ERG
  - Carries an ASPSCR1-TFE3 rather than YAP1-TFE3 fusion
  evidence:
  - reference: PMID:23737213
    reference_title: Novel YAP1-TFE3 fusion defines a distinct subset of epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The presence of diffuse expression of CD31 and/or ERG endothelial markers
      helps in the distinction from other TFE3-positive neoplasms, such as
      alveolar soft part sarcoma, PEComa, and Xp11-translocation positive renal
      cell carcinomas.
    explanation: >-
      Gives the immunophenotypic discriminator between YAP1-TFE3 EHE and other
      TFE3-driven tumours.

treatments:
- name: Complete Surgical Resection
  description: >-
    Complete resection is the mainstay and the only reliably curative option
    for localised, resectable EHE, and should be planned at a sarcoma
    reference centre. There is no established routine adjuvant systemic therapy
    or radiotherapy.
  therapeutic_modality: SURGERY
  treatment_term:
    preferred_term: Definitive Surgical Resection
    term:
      id: NCIT:C154430
      label: Definitive Surgical Resection
  evidence:
  - reference: PMID:34381186
    reference_title: "YAP1-TFE3-fused hemangioendothelioma: a multi-institutional clinicopathologic study of 24 genetically-confirmed cases."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Most patients (88%) were treated with primary surgical resection.
    explanation: >-
      Documents primary surgical resection as the dominant real-world management
      approach in a genetically confirmed cohort.
- name: Active Surveillance
  description: >-
    Because a substantial proportion of patients have stable, asymptomatic
    disease for years, an initial period of observation is often preferred to
    avoid the toxicity of systemic therapy. Objective progression, symptomatic
    deterioration, anaemia or new serosal involvement should prompt
    reconsideration of systemic treatment.
  treatment_term:
    preferred_term: active surveillance (watch-and-wait)
  notes: >-
    No NCIT clinical-action term for active surveillance is reachable from
    NCIT:C25218, so the term binding is intentionally omitted per the project's
    NCIT binding policy.
  evidence:
  - reference: PMID:29546487
    reference_title: "Epithelioid Hemangioendothelioma: Update on Diagnosis and Treatment."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Our treatment strategy is watch-and-wait versus active therapy on clinical
      trial or based on results of prior clinical trials.
    explanation: >-
      Expert-review statement of the watch-and-wait strategy. Evidence source is
      OTHER because this is a narrative treatment review.
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      One-third of these patients (n=18) were managed initially with expectant
      observation
    explanation: >-
      Quantifies real-world use of initial observation in a molecularly
      confirmed EHE cohort, showing this is a mainstream rather than
      exceptional approach.
- name: Sirolimus
  description: >-
    mTOR inhibition has the strongest recurring disease-specific activity
    signal in EHE. In a 38-patient retrospective series of progressive disease,
    sirolimus produced partial response in 10.8% and stable disease in 75.7%,
    with median progression-free survival of 13 months. Activity is markedly
    reduced in patients with serosal effusions. Use is off-label in most
    jurisdictions.
  therapeutic_modality: SMALL_MOLECULE
  target_mechanisms:
  - target: mTORC1 Pathway Activation
    treatment_effect: INHIBITS
    description: >-
      Sirolimus inhibits mTOR complex 1, the node on which fusion-driven
      YAP/TAZ activity converges.
  treatment_term:
    preferred_term: Pharmacotherapy
    term:
      id: NCIT:C15986
      label: Pharmacotherapy
    therapeutic_agent:
    - preferred_term: sirolimus
      term:
        id: CHEBI:9168
        label: sirolimus
  evidence:
  - reference: PMID:33107985
    reference_title: "Activity of sirolimus in patients with progressive epithelioid hemangioendothelioma: A case-series analysis within the Italian Rare Cancer Network."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The best RECIST responses were a partial response in 4 patients (10.8%),
      stable disease in 28 patients (75.7%), and disease progression in 5
      patients (13.5%).
    explanation: >-
      Quantifies the activity of sirolimus in progressive EHE.
  - reference: PMID:33107994
    reference_title: Sirolimus for patients with progressive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      The results from a robust retrospective cohort study support the use of
      sirolimus for patients who have multifocal epithelioid
      hemangioendothelioma without serosal effusion.
    explanation: >-
      Independent editorial appraisal endorsing sirolimus in effusion-negative
      multifocal EHE. Evidence source is OTHER because this is a commentary.
  - reference: PMID:39283723
    reference_title: GDF-15 Predicts Epithelioid Hemangioendothelioma Aggressiveness and Is Downregulated by Sirolimus through ATF4/ATF5 Suppression.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Sirolimus exhibited markedly higher antitumor activity compared with
      doxorubicin, concurrently reducing GDF-15 expression/release both in vivo
      and in vitro.
    explanation: >-
      Preclinical patient-derived model supporting sirolimus over anthracycline
      chemotherapy in EHE.
- name: Trametinib
  description: >-
    MEK inhibition is the mechanistically rationalised targeted approach,
    derived from the demonstration that TAZ-CAMTA1 signals through CTGF to the
    RAS-MAPK cascade. Trametinib abrogated fusion-driven anchorage-independent
    growth and xenograft expansion, which supported the SARC033 phase II trial
    in unresectable or metastatic EHE.
  therapeutic_modality: SMALL_MOLECULE
  target_mechanisms:
  - target: CTGF-Integrin-RAS-MAPK Signaling
    treatment_effect: INHIBITS
    description: >-
      Trametinib inhibits MEK downstream of the fusion-driven CTGF-RAS cascade,
      the node that sustains anchorage-independent growth.
  treatment_term:
    preferred_term: Targeted Therapy
    term:
      id: NCIT:C93352
      label: Targeted Therapy
    therapeutic_agent:
    - preferred_term: trametinib
      term:
        id: CHEBI:75998
        label: trametinib
  evidence:
  - reference: PMID:35443056
    reference_title: The TAZ-CAMTA1 Fusion Protein Promotes Tumorigenesis via Connective Tissue Growth Factor and Ras-MAPK Signaling in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Further, pharmacologic inhibition of MAPK signaling through PD 0325901 and
      trametinib abrogated TC-driven anchorage-independent growth.
    explanation: >-
      Preclinical basis for MEK inhibition as a fusion-directed therapy in EHE.
  - reference: PMID:35443056
    reference_title: The TAZ-CAMTA1 Fusion Protein Promotes Tumorigenesis via Connective Tissue Growth Factor and Ras-MAPK Signaling in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      These studies provided the preclinical rationale for SARC033 (NCI
      10015-NCT03148275), a nonrandomized, open-label, phase II study of
      trametinib in patients with unresectable or metastatic EHE.
    explanation: >-
      Links the preclinical mechanism directly to the clinical trial testing it.
- name: Investigational TAZ-TEAD Directed Therapy
  description: >-
    Because the fusion oncoprotein acts through TEAD and established tumours
    regress when it is switched off, direct disruption of the TAZ-CAMTA1-TEAD
    interaction is the most rational fusion-directed strategy. TEAD inhibitors
    have been tested against the first EHE cell lines, and a mechanistically
    distinct approach - restoring cytoplasmic sequestration of the fusion with
    a statin - abrogates fusion-driven anchorage-independent growth in
    endothelial models. Both remain preclinical/investigational with no
    clinical efficacy data in EHE.
  therapeutic_modality: SMALL_MOLECULE
  target_mechanisms:
  - target: Constitutive Nuclear TAZ-TEAD Transcriptional Activation
    treatment_effect: INHIBITS
    description: >-
      TEAD inhibition blocks the fusion's transcriptional output; statins act
      upstream by promoting its cytoplasmic sequestration.
  treatment_term:
    preferred_term: Targeted Therapy
    term:
      id: NCIT:C93352
      label: Targeted Therapy
  evidence:
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      EHE cell lines were established through ex vivo culture of tumor cells and
      evaluated for sensitivity to TEAD inhibition and trametinib.
    explanation: >-
      Documents TEAD inhibition as an actively tested fusion-directed strategy
      in the first EHE cell lines.
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Therefore, therapies that are known to inhibit YAP/TAZ activity also may
      inhibit TAZ-CAMTA1-induced endothelial cell transformation.
    explanation: >-
      Supports the broader class rationale, demonstrated with simvastatin
      abrogating soft-agar growth of TAZ-CAMTA1-expressing endothelial cells.
- name: Liver Transplantation
  description: >-
    Selected patients with unresectable but liver-confined hepatic EHE may be
    considered for transplantation at specialist centres. Selection is highly
    specialised and the supporting evidence is registry- and series-based.
  therapeutic_modality: SURGERY
  treatment_term:
    preferred_term: Liver Transplantation
    term:
      id: NCIT:C15271
      label: Liver Transplantation
  notes: >-
    No evidence item is attached: the transplantation literature for EHE is
    registry- and series-based and no quotable abstract statement meeting the
    project's snippet standard was identified during this curation pass. This
    is a deliberate, flagged gap rather than an oversight.
- name: Conventional Cytotoxic Chemotherapy
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Chemotherapy
    term:
      id: NCIT:C15632
      label: Chemotherapy
    therapeutic_agent:
    - preferred_term: doxorubicin
      term:
        id: CHEBI:28748
        label: doxorubicin
    - preferred_term: gemcitabine
      term:
        id: CHEBI:175901
        label: gemcitabine
    - preferred_term: paclitaxel
      term:
        id: CHEBI:45863
        label: paclitaxel
  description: >-
    Single- and multiagent doxorubicin-, gemcitabine-, and taxane-based regimens
    are used for systemic treatment of EHE, but there is no established standard
    cytotoxic regimen and the cited observational cohort does not establish
    controlled efficacy.
  notes: >-
    Paclitaxel represents the taxane class because it is specifically named in
    the separate EHE case series cited for paclitaxel plus bevacizumab below;
    PMID:31537895 reports taxane-based regimens at the class level. Common use in
    treatment records should not be interpreted as evidence of efficacy.
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Both single and multiagent cytotoxic regmens were also commonly used,
      either doxorubicin-, gemcitabine-, or taxane-based.
    explanation: >-
      Documents real-world use of these cytotoxic regimen classes. Support is
      PARTIAL because the observational cohort does not provide controlled
      evidence of efficacy.
- name: Local-Control Radiation Therapy
  therapeutic_modality: RADIOTHERAPY
  treatment_term:
    preferred_term: Radiation Therapy
    term:
      id: NCIT:C15313
      label: Radiation Therapy
  description: >-
    Radiation may be used as a local-control procedure alongside systemic
    treatment in selected EHE patients. This is distinct from routine adjuvant
    radiotherapy, which is not established after complete resection.
  notes: >-
    The cited cohort reports local modalities as a group and provides no
    controlled trial evidence for radiation-specific efficacy.
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      local treatment modalities, such as surgery, embolization, or radiation,
      were frequently utilized in conjunction with systemic therapy and often
      contributed to prolonged disease control.
    explanation: >-
      Identifies radiation among local-control modalities used with systemic
      therapy. Support is PARTIAL because this observational statement does not
      isolate radiation's effect or provide controlled efficacy evidence.
- name: Local-Control Embolization
  therapeutic_modality: SURGERY
  treatment_term:
    preferred_term: Embolization Therapy
    term:
      id: NCIT:C15230
      label: Embolization Therapy
  description: >-
    Embolization may be used as a local-control procedure alongside systemic
    treatment in selected EHE patients.
  notes: >-
    The cited cohort reports local modalities as a group and provides no
    controlled trial evidence for embolization-specific efficacy.
  evidence:
  - reference: PMID:31537895
    reference_title: Prognostic stratification of clinical and molecular epithelioid hemangioendothelioma subsets.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      local treatment modalities, such as surgery, embolization, or radiation,
      were frequently utilized in conjunction with systemic therapy and often
      contributed to prolonged disease control.
    explanation: >-
      Identifies embolization among local-control modalities used with systemic
      therapy. Support is PARTIAL because this observational statement does not
      isolate embolization's effect or provide controlled efficacy evidence.
- name: Paclitaxel plus Bevacizumab
  therapeutic_modality: MONOCLONAL_ANTIBODY
  treatment_term:
    preferred_term: Targeted Therapy
    term:
      id: NCIT:C93352
      label: Targeted Therapy
    therapeutic_agent:
    - preferred_term: bevacizumab
      term:
        id: NCIT:C2039
        label: Bevacizumab
    - preferred_term: paclitaxel
      term:
        id: CHEBI:45863
        label: paclitaxel
  description: >-
    Paclitaxel plus the anti-VEGF monoclonal antibody bevacizumab produced
    partial responses in a small hepatic EHE series, with prolonged
    stabilisation during bevacizumab maintenance in one patient. This is not an
    established standard of care.
  notes: >-
    The modality records the bevacizumab antibody platform, separating this
    combination from the pazopanib small-molecule arm; paclitaxel remains an
    explicit therapeutic agent. Bevacizumab is bound to NCIT rather than CHEBI
    because it is a monoclonal antibody.
  evidence:
  - reference: PMID:32201052
    reference_title: "VEGF-VEGFR pathway seems to be the best target in hepatic epithelioid hemangioendothelioma: A case series with review of the literature."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      While combination of paclitaxel and bevacizumab resulted in partial
      response in 3 patients, one of them also achieved long-term disease
      stabilization with bevacizumab maintenance with no adverse event.
    explanation: >-
      Reports the actual observed responses to VEGF-directed therapy. Support is
      PARTIAL because this is a four-patient case series, not a controlled
      trial.
- name: Pazopanib
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Targeted Therapy
    term:
      id: NCIT:C93352
      label: Targeted Therapy
    therapeutic_agent:
    - preferred_term: pazopanib
      term:
        id: CHEBI:71219
        label: pazopanib
  description: >-
    The oral multikinase inhibitor pazopanib produced clinical benefit in two
    patients in a small hepatic EHE series. It is not an established standard of
    care, and the supporting evidence is case-series level only.
  notes: >-
    Sorafenib is frequently mentioned in EHE review literature but is not
    supported by the case series cited here, so it is deliberately not listed
    as a `therapeutic_agent`.
  evidence:
  - reference: PMID:32201052
    reference_title: "VEGF-VEGFR pathway seems to be the best target in hepatic epithelioid hemangioendothelioma: A case series with review of the literature."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Two of the patients had clear benefit from pazopanib during the course of
      disease.
    explanation: >-
      Supports pazopanib activity in hepatic EHE at case-series level.

clinical_trials:
- name: NCT03148275
  phase: PHASE_II
  status: COMPLETED
  description: >-
    SARC033 (NCI 10015): non-randomised, open-label phase II study of the MEK
    inhibitor trametinib in patients with unresectable, locally advanced or
    metastatic epithelioid hemangioendothelioma. Built directly on the
    CTGF-RAS-MAPK dependence demonstrated preclinically.
  evidence:
  - reference: clinicaltrials:NCT03148275
    reference_title: "A Non-Randomized, Open-Label, Phase 2 Study of Trametinib in Patients With Unresectable or Metastatic Epithelioid Hemangioendothelioma"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This phase II trial studies how well trametinib works in treating patients
      with epithelioid hemangioendothelioma that has spread to other places in
      the body (metastatic), nearby tissue or lymph nodes (locally advanced), or
      cannot be removed by surgery (unresectable).
    explanation: >-
      Confirms the trial design and population for the fusion-directed MEK
      inhibition strategy.
- name: NCT06408441
  status: RECRUITING
  description: >-
    The EURACAN prospective clinical registry dedicated to EHE - a
    registry-based cohort study of newly diagnosed, molecularly confirmed EHE
    across 21 sarcoma reference centres in 10 European countries, designed to
    resolve the natural-history, prognostic-factor and treatment-efficacy gaps
    that dominate this disease.
  evidence:
  - reference: clinicaltrials:NCT06408441
    reference_title: "The Observational EURACAN Prospective Clinical Registry Dedicated to Epithelioid Hemangioendothelioma: the Protocol of an International and Collaborative Effort on an Ultra-rare Entity"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The objectives are to improve the understanding of EHE natural history,
      validate and identify new prognostic and predictive factors, clarify the
      activity and efficacy of currently available treatment options, describe
      treatment pattern.
    explanation: >-
      Documents the registry addressing the principal evidence gaps in EHE.

experimental_models:
- name: Conditional Wwtr1-Camta1 Knock-In Mouse
  description: >-
    The fusion is targeted to the endogenous Wwtr1 locus so that it is
    controlled by native transcriptional regulators upon Cre activation. Mice
    develop tumours indistinguishable from human EHE and no other tumour type,
    demonstrating both sufficiency and exquisite tissue specificity.
  evidence:
  - reference: PMID:33766982
    reference_title: WWTR1(TAZ)-CAMTA1 gene fusion is sufficient to dysregulate YAP/TAZ signaling and drive epithelioid hemangioendothelioma tumorigenesis.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Overall, these results demonstrate unequivocally that TAZ-CAMTA1 is
      sufficient to drive EHE formation with exquisite specificity, as no other
      tumor types were observed.
    explanation: >-
      Establishes the model's fidelity and the sufficiency of the fusion.
- name: Endothelial-Specific TAZ-CAMTA1 Mouse
  description: >-
    Expression of TAZ-CAMTA1 restricted to endothelial cells generates vascular
    tumours with the distinctive features of EHE; inhibiting the fusion causes
    regression, establishing continuing oncogene dependence.
  evidence:
  - reference: PMID:33766984
    reference_title: WWTR1(TAZ)-CAMTA1 reprograms endothelial cells to drive epithelioid hemangioendothelioma.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Our study provides the first genetic model of a TAZ fusion oncoprotein
      driving its associated human cancer, pinpointing TAZ-CAMTA1 as the key
      driver and a valid therapeutic target of EHE.
    explanation: >-
      Confirms the endothelial cell of origin and validates the fusion as a
      therapeutic target.
- name: EHE Cell Lines and Patient-Derived Xenograft
  description: >-
    The first EHE cell lines were generated from Cdkn2a-null TAZ-CAMTA1 mouse
    tumours and are addicted to the fusion oncoprotein; a separate 2024
    patient-derived xenograft with a matched cell line retains the
    WWTR1::CAMTA1 fusion and the donor tumour's histomorphology and
    genomic/transcriptomic profile.
  evidence:
  - reference: PMID:36598859
    reference_title: Loss of CDKN2A Cooperates with WWTR1(TAZ)-CAMTA1 Gene Fusion to Promote Tumor Progression in Epithelioid Hemangioendothelioma.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      These cell lines are "addicted" to the TC oncoprotein, replicate the EHE
      transcriptional profile, and generate EHE tumors when injected into
      immunodeficient mice.
    explanation: >-
      Documents oncogene addiction in the first EHE cell lines, the cellular
      basis for fusion-directed therapy.
  - reference: PMID:39283723
    reference_title: GDF-15 Predicts Epithelioid Hemangioendothelioma Aggressiveness and Is Downregulated by Sirolimus through ATF4/ATF5 Suppression.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      A patient-derived xenograft model and corresponding cell line were
      established from a patient with advanced EHE, demonstrating consistency
      with the original tumor in terms of histomorphology, WWTR1::CAMTA1 fusion
      presence, and genomic and transcriptomic profiles.
    explanation: >-
      Documents the human patient-derived model used for therapeutic and
      biomarker testing.

discussions:
- discussion_id: ehe_multifocality_clonality
  kind: KNOWLEDGE_GAP
  prompt: >-
    Does EHE multifocality arise by true metastatic dissemination from a single
    clone, or by field-effect / synchronous multicentric transformation?
  attaches_to:
  - "pathophysiology#Multifocal Dissemination and Serosal Involvement"
  rationale: >-
    Multifocality is present in about three quarters of cases at diagnosis and
    dominates management decisions. For multifocal *hepatic* disease the
    question is largely settled: separate lesions share an identical genomic
    breakpoint, i.e. clonal metastatic dissemination within one organ. What
    remains open is whether the same holds for the far commoner
    multi-compartment pattern (e.g. synchronous lung and liver, or lung and
    bone), where clonality has not been demonstrated at scale and where the
    supporting argument is still only circumstantial - outcome data showing no
    survival difference between multifocal and metastatic disease. The answer
    determines whether multi-organ multifocal EHE should be staged and treated
    as metastatic disease.
  proposed_experiments:
  - experiment_id: ehe_multiregion_breakpoint_sequencing
    name: Cross-compartment fusion-breakpoint sequencing of synchronous lesions
    description: >-
      Extend the published intrahepatic shared-breakpoint analysis to
      synchronous lesions in different compartments (lung plus liver, lung plus
      bone) using whole-genome or targeted breakpoint sequencing, to test
      whether multi-organ multifocality is also clonal rather than
      multicentric.
  - experiment_id: ehe_synchronous_lesion_phylogeny
    name: Phylogenetic reconstruction from secondary alterations
    description: >-
      Comparison of secondary alterations such as CDKN2A status across paired
      synchronous lesions to reconstruct clonal relationships.
- discussion_id: ehe_ctgf_mapk_model_fidelity
  kind: HUMAN_MODEL_MISMATCH
  prompt: >-
    Does the CTGF-integrin-RAS-MAPK arm mapped in TAZ-CAMTA1-transformed NIH3T3
    fibroblasts actually operate in human EHE endothelium?
  attaches_to:
  - "pathophysiology#CTGF-Integrin-RAS-MAPK Signaling"
  rationale: >-
    The entire CTGF/integrin/RAS-MAPK dependency - and with it the rationale for
    the SARC033 trametinib trial - was established in TAZ-CAMTA1-transformed
    NIH3T3 mouse fibroblasts and their xenografts. EHE is a tumour of
    endothelial lineage, and the two independent genetic mouse models of the
    disease both act in the endothelial compartment. Whether a fibroblast-based
    transformation assay faithfully reports the signalling wiring of human EHE
    endothelium is unresolved, and this is precisely the kind of mismatch that
    can make a mechanistically compelling target fail clinically. This is a
    HUMAN_MODEL_MISMATCH rather than a KNOWLEDGE_GAP because the evidence
    exists and is strong - it is its translational validity that is in
    question.
  proposed_experiments:
  - experiment_id: ehe_ctgf_mapk_in_endothelial_models
    name: Re-test CTGF-RAS-MAPK dependence in endothelial and patient-derived EHE models
    description: >-
      Repeat the CTGF knockdown, integrin-blockade and MEK-inhibition
      experiments in TAZ-CAMTA1-expressing endothelial cells, the established
      EHE cell lines and the patient-derived xenograft, rather than in NIH3T3
      fibroblasts.
  - experiment_id: ehe_mapk_activity_in_human_tumours
    name: MAPK pathway activity in human EHE tissue
    description: >-
      Quantify phospho-ERK and CTGF expression in primary human EHE specimens,
      and correlate with trametinib response among SARC033 participants, to
      confirm the pathway is active in the human disease.
- discussion_id: ehe_serosal_effusion_sirolimus_resistance
  kind: KNOWLEDGE_GAP
  prompt: >-
    Why do patients with serosal effusion respond so poorly to sirolimus, and
    is there a distinct biology underlying effusion-positive EHE?
  attaches_to:
  - "pathophysiology#Multifocal Dissemination and Serosal Involvement"
  rationale: >-
    Serosal effusion is the most reproducible adverse prognostic factor in EHE
    and specifically predicts sirolimus failure, but no molecular correlate has
    been identified. Whether effusion marks a distinct tumour biology, a
    pharmacokinetic barrier, or simply advanced disease burden is unresolved,
    and the answer determines whether a different therapeutic class is needed
    for this subgroup.
  proposed_experiments:
  - experiment_id: ehe_effusion_comparative_profiling
    name: Comparative profiling of effusion-positive versus effusion-negative EHE
    description: >-
      Transcriptomic and secondary-mutation profiling of matched
      effusion-positive and effusion-negative tumours to identify a molecular
      correlate of the adverse phenotype.
  - experiment_id: ehe_effusion_mtorc1_pharmacodynamics
    name: Pharmacodynamic assessment of mTORC1 inhibition in effusion-positive disease
    description: >-
      Measurement of intratumoural sirolimus exposure and phospho-4E-BP1 as a
      pharmacodynamic readout, to distinguish drug-delivery failure from
      target-independent resistance.
- discussion_id: ehe_yap1_tfe3_nosology
  kind: KNOWLEDGE_GAP
  prompt: >-
    Should YAP1-TFE3-fused hemangioendothelioma be classified as a separate
    disease entity rather than an EHE subtype?
  attaches_to:
  - "pathophysiology#YAP1-TFE3 Fusion Oncogene Formation"
  rationale: >-
    The YAP1-TFE3 subset differs from conventional EHE in driver fusion,
    architecture, immunophenotype, patient age and survival, and expert groups
    have proposed splitting it out. MONDO:0015523 currently subsumes both. The
    nosology decision affects trial eligibility and prognostic communication,
    and turns on whether the two converge on a common downstream transcriptional
    program.
  proposed_experiments:
  - experiment_id: ehe_tead_vs_mit_chromatin_profiling
    name: Comparative TEAD versus MiT target chromatin profiling
    description: >-
      Direct comparative transcriptomic and chromatin-occupancy profiling of
      WWTR1-CAMTA1 versus YAP1-TFE3 tumours to determine whether the two fusions
      converge on a shared downstream program.
  - experiment_id: ehe_fusion_stratified_outcome_analysis
    name: Fusion-stratified prospective outcome analysis
    description: >-
      Pooled prospective outcome analysis stratified by fusion; the EURACAN
      registry (NCT06408441) is positioned to deliver this.

classifications:
  icdo_morphology:
    classification_value: Sarcoma
  harrisons_chapter:
  - classification_value: ONCOLOGY_HEMATOLOGY

📚

References & Deep Research

Deep Research

1

Falcon ▸

Epithelioid Hemangioendothelioma: Comprehensive Disease-Characteristics Report

Edison Scientific Literature 18 citations 2026-08-01T17:14:47.116071

Epithelioid Hemangioendothelioma: Comprehensive Disease-Characteristics Report

Evidence scope. This synthesis prioritizes molecularly confirmed human disease, expert-center registry/consensus material, and primary mechanistic studies, especially 2022–2024. Older numerical estimates are retained when newer prospective data are unavailable. Most information is aggregated disease-level evidence rather than patient-level EHR data. The prospective EURACAN registry begun in December 2023 is designed to remedy major natural-history and treatment-evidence gaps across 21 expert centers in 10 countries. (NCT06408441 chunk 1)

Domain	Key facts	Ontology / coding suggestions
Identity / classification	Epithelioid hemangioendothelioma (EHE) is an ultra-rare malignant vascular sarcoma / soft-tissue tumor of endothelial lineage; often described as a translocation-associated, YAP/TAZ-driven cancer with highly variable clinical behavior (stacchiotti2024gdf15predictsepithelioid pages 1-2, seavey2022unravelingthebiology pages 2-4, lamar2018epithelioidhemangioendotheliomaas pages 1-3)	MONDO: epithelioid hemangioendothelioma = MONDO:0015523 (from available disease-target context; verify locally). MeSH / ICD / OMIM / Orphanet: verify locally. NCIT: verify locally.
Epidemiology	Estimated prevalence is less than 1 per 1,000,000; EHE accounts for <1% of vascular tumors. Median/typical age at presentation is around 35-40 years, with broad reported range from 8-90 years. Female predominance has been reported (~1.5:1) (seavey2022unravelingthebiology pages 2-4, lamar2018epithelioidhemangioendotheliomaas pages 1-3)	Rare cancer category; disease-level aggregated literature and registry data, not individual EHR-derived.
Main sites	Most frequent sites include liver, lung, bone, and soft tissue; one review summarized liver 21%, bone 14%, lung 12%. Multifocal liver disease is a common presentation; tumors can arise in many organs/tissues (lamar2018epithelioidhemangioendotheliomaas pages 5-7, lamar2018epithelioidhemangioendotheliomaas pages 1-3)	UBERON suggestions: liver (UBERON:0002107), lung (UBERON:0002048), bone tissue (UBERON:0002481), soft tissue = verify locally. Cell type: endothelial cell (CL:0000115).
Defining alterations	Approx. 90% harbor WWTR1(TAZ)::CAMTA1 fusion; the remaining ~10% harbor YAP1::TFE3 fusion. In one synthesis, >90% carried WWTR1::CAMTA1 and ~45% of WWTR1::CAMTA1-positive tumors had no other genetic alterations. Secondary alterations reported in ~50% include CDKN2A/B, RB, APC, ATRX, XRCC2, FANCA (stacchiotti2024gdf15predictsepithelioid pages 1-2, seavey2022unravelingthebiology pages 2-4)	HGNC genes: WWTR1, CAMTA1, YAP1, TFE3, CDKN2A, CDKN2B, RB1, APC, ATRX, XRCC2, FANCA. Structural variant class: gene fusion / chromosomal translocation.
Mechanism	Fusion oncoproteins preserve TEAD-binding function and drive constitutive Hippo-pathway effector activity, producing TEAD-dependent transcription, altered chromatin regulation via the ATAC histone acetyltransferase complex, and oncogenic programs promoting proliferation, survival, migration, anchorage-independent growth, and metastasis. TAZ-CAMTA1 also signals through CTGF, integrin αIIbβ3, and Ras-MAPK; MEK inhibition suppresses fusion-driven growth. CAMTA1 contributes a nuclear-localization function that helps evade normal Hippo cytoplasmic restraint (lamar2018epithelioidhemangioendotheliomaas pages 3-5, seavey2022unravelingthebiology pages 7-8, stacchiotti2024gdf15predictsepithelioid pages 1-2)	GO suggestions: Hippo signaling (verify locally), regulation of transcription by RNA polymerase II (GO:0006357), chromatin organization (GO:0006325), cell proliferation (GO:0008283), cell migration (GO:0016477), anoikis / apoptotic process = verify locally. CL: endothelial cell (CL:0000115).
Diagnostic hallmarks	Diagnosis integrates morphology, endothelial differentiation, and molecular testing. WWTR1::CAMTA1 is pathognomonic/standard molecular marker; CAMTA1 immunohistochemistry can be a useful surrogate marker. YAP1::TFE3 defines a distinct subset, but TFE3 immunohistochemistry is not fully specific. Histology typically shows epithelioid endothelial cells in a prominent fibrous/myxohyaline stroma (ong2021genefusionsin pages 13-13, lamar2018epithelioidhemangioendotheliomaas pages 5-7, seavey2022unravelingthebiology pages 2-4)	Diagnostic workflow: histopathology + endothelial immunophenotype + fusion confirmation by RNA sequencing/FISH/targeted molecular assay (verify locally for assay standards). HPO / SNOMED / LOINC coding: verify locally.
Prognosis	Clinical course is markedly heterogeneous, from indolent disease lasting years to rapidly fatal progression within months. Review-level survival figures include median overall survival ~75 months and 1-, 3-, 5-year survival of 83.4%, 55.7%, and 41.1%, respectively. Worse outcomes are associated with pleural/peritoneal disease; isolated soft-tissue disease has more favorable 5-year survival (~87%), compared with liver (~65%) and lung (~45%). Tumor size >3 cm and >3 mitoses/HPF were associated with worse 5-year survival (59% vs 100% in one review synthesis) (lamar2018epithelioidhemangioendotheliomaas pages 3-5, seavey2022unravelingthebiology pages 2-4, lamar2018epithelioidhemangioendotheliomaas pages 1-3)	Prognostic factor concepts: pleural effusion / serosal involvement, tumor size, mitotic activity. HPO suggestions: pleural effusion (HP:0002202), ascites (HP:0001541), pain (HP:0012531), weight loss (HP:0001824).
Treatment	For localized disease, surgery is the mainstay and can be curative for solitary lesions. Conventional chemotherapy is generally considered less effective in EHE than in angiosarcoma. Targeted/systemic approaches under study or use include sirolimus, trametinib, pazopanib, sorafenib, eribulin, and local ablative/interventional strategies in selected hepatic disease. In preclinical comparison, sirolimus showed greater antitumor activity than doxorubicin; in older case-series summaries, sirolimus showed prolonged benefit in a subset of patients (stacchiotti2024gdf15predictsepithelioid pages 1-2, stacchiotti2024gdf15predictsepithelioid pages 9-9, lamar2018epithelioidhemangioendotheliomaas pages 5-7, lamar2018epithelioidhemangioendotheliomaas pages 7-8)	NCIT intervention suggestions: Surgical Resection, Sirolimus, Trametinib, Pazopanib, Sorafenib, Eribulin, Doxorubicin, Liver Transplantation, Ablation (verify locally for exact NCIT IDs). Clinical trials: trametinib NCT03148275; eribulin NCT03331250; nab-sirolimus NCT07104331; albumin-bound sirolimus NCT07684287 (NCT03148275 chunk 3).
Emerging biomarker	Circulating GDF-15 is a 2024 candidate biomarker of aggressiveness and treatment monitoring. In two cohorts, GDF-15 was significantly higher in EHE than controls (P<0.001 in both), highest in higher-risk disease (P=0.006 retrospective; P=0.002 prospective), and correlated with aggressiveness. Example longitudinal values rose from 6,792 to 11,475 pg/mL with progression and fell to 4,064 pg/mL with sirolimus-induced stability (stacchiotti2024gdf15predictsepithelioid pages 11-12)	CHEBI / protein identifier coding: GDF-15 verify locally. Biomarker type: circulating protein biomarker.
Models	Strong model evidence now exists: (1) genetically engineered conditional Wwtr1-Camta1 mouse models showing that the fusion is sufficient to drive bona fide EHE; (2) Tet-Off overexpression mouse systems generating postnatal lesions; (3) NIH3T3 transformed-cell and xenograft models for mechanistic work; (4) 2024 patient-derived xenograft and matched cell line retaining WWTR1::CAMTA1 and tumor genomic/transcriptomic features (seavey2022unravelingthebiology pages 7-8, stacchiotti2024gdf15predictsepithelioid pages 1-2)	Evidence types: in vitro, xenograft, genetically engineered mouse, patient-derived xenograft. Model organism: mouse (NCBI Taxon:10090).
Prevention	No established primary prevention, population screening, or protective-factor evidence is currently supported. EHE is largely fusion-driven, usually somatic, with no validated inherited predisposition pattern or confirmed environmental cause in the cited evidence. Prevention emphasis is therefore tertiary: expert pathology review, accurate fusion testing, surveillance of known disease, and management of complications (stacchiotti2024gdf15predictsepithelioid pages 1-2, lamar2018epithelioidhemangioendotheliomaas pages 1-3)	Prevention coding: verify locally. Genetic counseling may be considered case-by-case, but routine hereditary screening is not evidence-based from available sources.

Table: This table condenses the highest-yield disease characteristics of epithelioid hemangioendothelioma for knowledge-base use, including defining fusions, mechanism, prognosis, treatment, and emerging biomarker evidence. It highlights exact reported percentages and flags ontology or coding elements that should be verified locally before database ingestion.

1. Disease information

Definition

Epithelioid hemangioendothelioma (EHE) is an ultra-rare, translocation-associated malignant vascular sarcoma showing endothelial differentiation. It can arise almost anywhere, but liver, lung, bone, and soft tissue are especially frequent. Its behavior ranges from prolonged radiographic stability or spontaneous regression to rapidly fatal metastatic disease. (stacchiotti2024gdf15predictsepithelioid pages 1-2, NCT06408441 chunk 1)

Identifiers and terminology

MONDO: MONDO:0015523.
Common names: epithelioid hemangioendothelioma; epithelioid haemangioendothelioma; EHE; hepatic EHE/HEHE; pulmonary EHE/PEH when site-qualified.
Historical term: intravascular bronchioloalveolar tumor, formerly used for pulmonary EHE.
MeSH: Hemangioendothelioma, Epithelioid; exact current identifier should be verified against the release used by the knowledge base.
Orphanet, OMIM, ICD-10/ICD-11: release-specific codes were not reliably present in the retrieved primary literature and should be resolved directly against current terminology services. EHE is a somatic neoplasm, not a classic Mendelian phenotype; an OMIM disease entry should not be assumed.
Coding caution: ICD-10-CM often cannot represent EHE histology independently of anatomical site. Cancer-registry coding should combine topography with the applicable ICD-O morphology code after verification against the current ICD-O release.

Open Targets recognizes EHE as MONDO:0015523 but returned no disease-specific target associations; angiosarcoma associations must not be transferred to EHE because these are biologically distinct entities. (OpenTargets Search: epithelioid hemangioendothelioma)

2. Etiology, risk, and protective factors

Causal lesion

EHE is principally caused by an acquired somatic structural rearrangement, not by inherited pathogenic variation:

Approximately 90% of tumors harbor WWTR1::CAMTA1, generated classically by t(1;3)(p36;q25).
A minority historically classified as EHE harbor YAP1::TFE3. Some contemporary experts recommend calling these YAP1::TFE3-fused hemangioendothelioma because their morphology and biology differ from classic WWTR1::CAMTA1-positive EHE. (stacchiotti2024gdf15predictsepithelioid pages 1-2, ma2022thetazcamta1fusion pages 1-2)

The reported proportions vary with classification and assay—approximately 80–>90% for WWTR1::CAMTA1 and roughly 10–12% for YAP1::TFE3—so frequencies should be interpreted in the context of diagnostic criteria. (ong2021genefusionsin pages 8-10, stacchiotti2024gdf15predictsepithelioid pages 1-2)

Risk and protective factors

No validated inherited susceptibility allele, founder mutation, carrier state, family-history effect, infectious trigger, occupational exposure, toxin, smoking/alcohol association, dietary determinant, or protective lifestyle factor has been established. The recognized arsenic, vinyl chloride, thorium dioxide, radiation, and chronic-lymphedema risks for angiosarcoma should not be attributed to EHE. No reproducible gene–environment interaction is known. Accordingly, penetrance, anticipation, germline mosaicism, consanguinity, and carrier frequency are not applicable to the usual somatic EHE mechanism.

3. Phenotypes and quality-of-life effects

EHE usually begins in adulthood, around the mid-30s to 40 years, but the reported range extends from childhood to advanced age (approximately 8–90 years). About 25–33% of patients may be asymptomatic and diagnosed incidentally. (seavey2022unravelingthebiology pages 2-4, lamar2018epithelioidhemangioendotheliomaas pages 1-3)

Phenotype	Type, course, and impact	Suggested HPO term
Pain, including tumor or bone pain	Symptom; site-dependent, variable; can impair mobility, sleep, and work and may trigger systemic treatment	Pain, HP:0012531; Bone pain, HP:0002653
Weight loss/anorexia/fatigue	Constitutional symptoms; associated systemic deterioration is concerning for aggressive disease	Weight loss, HP:0001824; Feeding/appetite abnormality and Fatigue—verify release
Cough, dyspnea, chest pain	Pulmonary/pleural manifestations; severity ranges from mild to respiratory compromise	Cough, HP:0012735; Dyspnea, HP:0002094; Chest pain, HP:0100749
Pleural or peritoneal effusion/ascites	Clinical/imaging sign; progressive serosal disease is a major adverse feature	Pleural effusion, HP:0002202; Ascites, HP:0001541
Hepatomegaly/right-upper-quadrant discomfort	Hepatic manifestation; may remain indolent or progress to portal/hepatic dysfunction	Hepatomegaly, HP:0002240; Abdominal pain, HP:0002027
Pathologic fracture or impaired mobility	Bone disease; uncommon overall but potentially severe	Pathologic fracture, HP:0002756; Abnormality of mobility—verify release
Anemia	Laboratory abnormality; prospectively unvalidated but retrospectively associated with worse advanced disease	Anemia, HP:0001903
Multifocal/metastatic lesions	Imaging/pathologic phenotype rather than symptom; lung, liver, and bone are common compartments	Neoplasm/metastatic disease terms should be drawn from NCIT/SNOMED rather than forced into HPO

The NCI trametinib study formally incorporated PROMIS global health, pain intensity, pain interference, and pain behavior, reflecting clinically important quality-of-life burdens. However, disease-specific validated EHE quality-of-life norms remain limited. (NCT03148275 chunk 1)

4. Genetic and molecular information

Principal alterations

WWTR1 encodes TAZ, a Hippo-pathway transcriptional coactivator; CAMTA1 supplies nuclear localization and transcription/chromatin-regulatory functions. The fusion is a somatic, activating structural variant rather than a conventional ACMG germline SNV. It is therefore best classified as an oncogenic/pathogenic somatic fusion rather than assigned a germline ACMG category.
YAP1::TFE3 is another activating somatic fusion involving the YAP Hippo effector.
Population allele frequency in gnomAD/1000 Genomes is not meaningful for tumor-specific chromosomal fusions; they are expected to be absent from normal-population germline datasets.
Secondary alterations occur in about half of tumors in some series and include CDKN2A/CDKN2B, RB1, APC, ATRX, XRCC2, and FANCA. Approximately 45% of WWTR1::CAMTA1-positive tumors in one synthesis had no other detected alteration, underscoring the fusion’s sufficiency. (seavey2022unravelingthebiology pages 7-8, stacchiotti2024gdf15predictsepithelioid pages 1-2)

No clinically validated modifier gene is established. Loss of cell-cycle restraints such as CDKN2A is biologically plausible as a progression event, but currently does not define routine genotype-guided care.

Epigenetic/chromatin biology

Both fusion proteins recruit the Ada2a-containing histone acetyltransferase (ATAC) complex, including YEATS2 and ZZZ3, while hyperactivating TEAD-dependent transcription. This is functional chromatin dysregulation rather than a validated diagnostic DNA-methylation signature. ATAC-complex depletion inhibits anchorage-independent growth in fusion-transformed cells, making the complex a candidate therapeutic vulnerability. (seavey2022unravelingthebiology pages 7-8, stacchiotti2024gdf15predictsepithelioid pages 1-2)

5. Environmental and infectious information

No specific environmental, lifestyle, occupational, radiation, medication, or infectious cause has been demonstrated for EHE. Apparent hormonal or female-sex associations remain descriptive and do not establish causation. No pathogen is implicated, and the disease is neither communicable nor zoonotic.

6. Mechanism and pathophysiology

Causal chain

Somatic rearrangement → fusion transcriptional coactivator → constitutive nuclear YAP/TAZ-like activity → TEAD-dependent transcription plus ATAC-mediated chromatin remodeling → survival, proliferation, migration, anoikis resistance, fibromyxoid-stroma production, and metastatic competence → organ-specific nodules, pain, effusions, and organ dysfunction.

Mechanistically, TAZ–CAMTA1 retains the TAZ TEAD-binding region, loses normal C-terminal regulatory elements, and gains CAMTA1-mediated nuclear localization. Consequently, normal Hippo/LATS phosphorylation does not efficiently impose cytoplasmic restraint. (lamar2018epithelioidhemangioendotheliomaas pages 3-5, ma2022thetazcamta1fusion pages 1-2)

A second experimentally supported branch is:

TAZ–CAMTA1 → CTGF induction → CTGF binding integrin αIIbβ3 → RAS–MAPK activation → anchorage-independent growth and xenograft expansion. CTGF knockdown or MEK inhibition with PD0325901/trametinib suppressed growth in transformed NIH3T3 cultures and xenografts. This provided the rationale for phase II trametinib trial NCT03148275. The article’s abstract states: “CTGF and the Ras-MAPK signaling cascade are essential for TC-mediated tumorigenesis.” (PMID 35443056; published April 2022; DOI: https://doi.org/10.1158/1078-0432.CCR-22-0421). (ma2022thetazcamta1fusion pages 1-2)

mTOR and GDF-15

The 2024 study identified tumor-derived GDF-15 as a candidate aggressiveness biomarker. Sirolimus inhibited 4E-BP1 phosphorylation, reduced ATF4/ATF5, and lowered GDF-15 expression/release. In two human cohorts, GDF-15 was higher in EHE than controls (both P<0.001), highest in higher-risk disease (P=0.006 retrospective; P=0.002 prospective), and correlated with aggressiveness. In one longitudinal example, levels rose from 6,792 to 11,475 pg/mL during progression and fell to 4,064 pg/mL with sirolimus-associated stability. This remains a candidate—not yet validated or regulatory-qualified—biomarker. (stacchiotti2024gdf15predictsepithelioid pages 9-9, stacchiotti2024gdf15predictsepithelioid pages 11-12)

The 2024 abstract states: “This study identifies GDF-15 as a novel biomarker of EHE aggressiveness” and reports “markedly higher antitumor activity” for sirolimus than doxorubicin in experimental models (published September 2024; DOI: https://doi.org/10.1158/1078-0432.CCR-23-3991). (stacchiotti2024gdf15predictsepithelioid pages 1-2)

Molecular-profile and ontology suggestions

GO biological processes: Hippo signaling; regulation of transcription by RNA polymerase II (GO:0006357); chromatin organization (GO:0006325); histone acetylation (GO:0016573); cell proliferation (GO:0008283); cell migration (GO:0016477); MAPK cascade (GO:0000165); angiogenesis (GO:0001525); apoptotic signaling/anoikis—verify current term.
Cell Ontology: endothelial cell (CL:0000115); vascular endothelial cell and organ-specific endothelial subtypes should be added when supported by specimen location.
GO cellular components: nucleus (GO:0005634), chromatin (GO:0000785), transcription-regulator complex (GO:0005667).
Robust EHE metabolomic, lipidomic, spatial-transcriptomic, and single-cell atlases are not yet established. Current multi-omic evidence is primarily bulk genomic/transcriptomic and proteomic/chromatin-interaction work.

7. Anatomy

Primary or multifocal disease most often affects liver (UBERON:0002107), lung (UBERON:0002048), bone tissue (UBERON:0002481), and soft/connective tissue. Pleura, peritoneum, mediastinum, skin, central nervous system, head and neck, and other viscera can also be involved. One literature synthesis reported liver 21%, bone 14%, and lung 12%, but these estimates are referral- and classification-sensitive. (lamar2018epithelioidhemangioendotheliomaas pages 1-3)

EHE arises from endothelial-lineage cells within connective/vascular tissue. At the subcellular level, the defining dysfunction is nuclear localization of fusion coactivators and altered chromatin/transcriptional complexes. Disease may be unifocal, multifocal within one organ/compartment, or systemic; there is no characteristic lateralization. (NCT06408441 chunk 1)

8. Temporal development and natural history

Onset is generally insidious. There is no universal AJCC stage system specific to EHE; practical grouping is localized/unifocal, locoregional or multifocal single-compartment, and systemic metastatic disease. The disease may remain stable untreated, slowly progress over years, spontaneously regress rarely, or become rapidly fatal. (NCT06408441 chunk 1)

Because a radiographic change can be slow and conventional therapy has toxicity, experts often use an initial observation interval for asymptomatic advanced disease. Symptomatic deterioration, objective progression, anemia, or new serosal involvement/effusion constitutes an important window to reconsider systemic therapy. Prospective validation is ongoing in EURACAN’s registry, which mandates six-monthly updates when no event occurs. (NCT06408441 chunk 1)

9. Inheritance and population epidemiology

The best recent disease-specific estimate cited by the EURACAN registry is an incidence of 0.038 per 100,000 person-years—approximately 0.38 per million/year. Prevalence has been described as below one per million, although estimates are uncertain because of historical misclassification and long survival in indolent cases. (wiegand2021malignantvasculartumors pages 7-9, NCT06408441 chunk 1)

Typical age is approximately 35–40 years, with a wide 8–90-year range. A female predominance around 1.5:1 has been reported, although site-specific cohorts differ. No reproducible ethnic, geographic, or founder distribution is established. (seavey2022unravelingthebiology pages 2-4, lamar2018epithelioidhemangioendotheliomaas pages 1-3)

EHE is not ordinarily inherited: no autosomal-dominant, recessive, X-linked, mitochondrial, or polygenic pattern is established. Thus penetrance, carrier frequency, prenatal diagnosis, cascade testing, and preimplantation testing are generally not applicable.

10. Diagnostics

Recommended workflow

Imaging and staging: contrast-enhanced CT of chest/abdomen/pelvis; liver MRI for hepatic disease; MRI for soft-tissue/bone disease; targeted bone imaging or PET/CT when clinically indicated. Pulmonary EHE often presents as multiple nodules and/or pleural thickening. Hepatic EHE commonly presents as multifocal peripheral nodules, sometimes calcified, with peripheral enhancement and the radiologic “lollipop sign.” MRI generally shows low/intermediate T1 and high T2 signal with enhancement. (lamar2018epithelioidhemangioendotheliomaas pages 3-5, wiegand2021malignantvasculartumors pages 7-9)
Core biopsy/excision reviewed by an expert sarcoma pathologist. Histology typically shows cords, strands, or nests of epithelioid endothelial cells in dense myxohyaline stroma; intracytoplasmic lumina/vacuoles may represent primitive vascular differentiation. (seavey2022unravelingthebiology pages 2-4)
Immunohistochemistry: CD31/PECAM1 and ERG are strong endothelial markers; CD34 and FLI1 are commonly supportive. Nuclear CAMTA1 is a useful surrogate for WWTR1::CAMTA1. TFE3 staining is insufficiently specific and should be molecularly verified. (lamar2018epithelioidhemangioendotheliomaas pages 3-5, wiegand2021malignantvasculartumors pages 7-9, ong2021genefusionsin pages 8-10)
Molecular confirmation: fusion FISH, targeted RNA sequencing, anchored multiplex RNA assay, RT-PCR when breakpoints/transcripts are covered, or whole-transcriptome sequencing. A negative CAMTA1/FISH assay in morphologically convincing disease should prompt an RNA-based fusion assay rather than automatically excluding EHE. (seavey2022unravelingthebiology pages 2-4)

The NCI trametinib trial required histologically confirmed disease and tissue for fusion FISH, illustrating real-world integration of molecular confirmation into prospective research. (NCT03148275 chunk 2)

Differential diagnosis

Important alternatives are epithelioid angiosarcoma, epithelioid hemangioma, pseudomyogenic hemangioendothelioma, metastatic carcinoma, melanoma, myoepithelial neoplasm, and other epithelioid sarcomas. Marked atypia, destructive vasoformation, brisk mitoses, and necrosis favor angiosarcoma; keratins alone cannot establish carcinoma because vascular tumors may show focal epithelial-marker expression. Fusion confirmation is decisive in difficult cases. (lamar2018epithelioidhemangioendotheliomaas pages 3-5, seavey2022unravelingthebiology pages 2-4)

Routine blood tests are nonspecific; CBC, liver/renal function, inflammatory indices, hemoglobin, and fibrinogen are useful for burden, safety, and longitudinal assessment. GDF-15, CTGF, CRP, and ESR remain investigational. There is no population, newborn, carrier, or hereditary-family screening program. Liquid biopsy, WES/WGS, CMA, karyotyping, mitochondrial testing, and repeat-expansion testing are not routine first-line diagnostics; RNA-based fusion detection is generally more directly informative.

11. Outcome and prognosis

Clinical outcome is heterogeneous. Review-level estimates include median overall survival around 75 months and 1-, 3-, and 5-year survival of 83.4%, 55.7%, and 41.1%, respectively; these pooled historical figures should not be treated as individualized predictions. Reported 5-year survival differs by compartment—approximately 87% for isolated soft tissue, 65% with liver involvement, and 45% with lung involvement. (lamar2018epithelioidhemangioendotheliomaas pages 3-5, lamar2018epithelioidhemangioendotheliomaas pages 5-7, seavey2022unravelingthebiology pages 2-4)

In one risk synthesis, tumors >3 cm and >3 mitoses per high-power field were associated with 5-year survival of 59%, versus 100% without those features. Recurrence after treatment was approximately 15%, and regional metastasis 20–30%, although estimates vary by site and cohort. (lamar2018epithelioidhemangioendotheliomaas pages 3-5)

The most clinically persuasive adverse pattern is serosal/pleural or peritoneal involvement, especially effusion, systemic symptom deterioration, and anemia; prospective confirmation remains pending. Other reported adverse associations include older age, male sex, pulmonary or multiorgan disease, larger size, and higher mitotic activity. (wiegand2021malignantvasculartumors pages 7-9, NCT06408441 chunk 1)

Morbidity includes chronic pain, dyspnea, effusions, impaired mobility/fracture, liver dysfunction, repeated imaging and procedures, treatment toxicity, and uncertainty from an unpredictable course. Recovery is plausible after complete local treatment, but disseminated disease is usually chronic and requires long-term surveillance.

12. Treatment and real-world implementation

Strategy

Localized resectable disease: complete surgery is preferred; expected cure is approximately 70–80% in the EURACAN protocol summary. There is no proven routine adjuvant systemic therapy or radiotherapy. (NCT06408441 chunk 1)
Asymptomatic stable advanced disease: active surveillance is often preferred to avoid overtreatment.
Progressive, symptomatic, organ-threatening, or serosa-involving disease: systemic treatment and/or site-directed intervention should be decided by a multidisciplinary sarcoma center.
Hepatic disease: resection, transplantation, ablation, or embolization may be considered in selected patients; extrahepatic disease does not automatically preclude transplantation in all historical series, but selection is highly specialized. (lamar2018epithelioidhemangioendotheliomaas pages 5-7)
Radiotherapy: may provide local control or palliation in selected unresectable, painful, or osseous lesions; evidence is retrospective.

Systemic therapies

Sirolimus/other mTOR inhibition. Sirolimus has the strongest recurrent disease-specific activity signal and is often regarded by experts as a preferred systemic option for progressive EHE, particularly without severe serosal effusion. It is off-label in many jurisdictions. The 2024 PDX study found an in-vitro IC50 of 0.03 ± 0.03 μmol/L for sirolimus versus 0.10 ± 0.04 μmol/L for doxorubicin and superior experimental antitumor activity; this comparison is preclinical and cannot establish clinical superiority. (stacchiotti2024gdf15predictsepithelioid pages 9-9, stacchiotti2024gdf15predictsepithelioid pages 1-2)

Trametinib. The mechanistic rationale is CTGF–RAS–MAPK dependence. NCT03148275/SARC033 was a completed, single-arm phase II trial of oral trametinib in 44 patients with unresectable/metastatic EHE; primary outcome was RECIST 1.1 response, with PFS, OS, toxicity, PROMIS measures, CTGF, inflammatory markers, fusion status, and MAPK activity assessed. Registry results were first posted November 26, 2024; because exact efficacy values were not captured in the retrieved evidence, they are not inferred here. (NCT03148275 chunk 1, NCT03148275 chunk 2)

Conventional chemotherapy. Anthracycline-based and other cytotoxic regimens generally show marginal activity and are not interchangeable with angiosarcoma treatment. Doxorubicin, taxanes, carboplatin, and other sarcoma regimens may still be considered in exceptional rapidly progressive cases, but expectations should be modest. (lamar2018epithelioidhemangioendotheliomaas pages 7-8, NCT06408441 chunk 1)

Antiangiogenic/targeted agents. Pazopanib and sorafenib have produced prolonged stable disease or occasional responses in reports and small series, but no universally accepted standard exists. Immunotherapy, gene therapy, RNA therapy, CAR-T/cell therapy, and fusion-directed editing remain experimental without established efficacy.

Eribulin. NCT03331250 is a pilot, open-label phase II study of 13 total angiosarcoma/EHE participants, with no more than five EHE patients planned. Eribulin inhibits microtubule function; disease-specific conclusions are necessarily limited by the tiny EHE subgroup. (NCT03331250 chunk 1)

Current research infrastructure. The prospective EURACAN registry (NCT06408441) began enrolling molecularly confirmed new cases in December 2023, targeting 100 participants and longitudinally recording symptoms, pain, blood tests, serosal disease, imaging, surveillance, and treatment outcomes. (NCT06408441 chunk 1)

Suggested NCIT intervention concepts include Surgical Resection, Active Surveillance, Liver Transplantation, Radiation Therapy, Ablation, Sirolimus, Trametinib, Pazopanib, Sorafenib, Eribulin, and Doxorubicin; exact NCIT identifiers should be resolved against the current thesaurus. Pharmacogenomic dosing rules specific to EHE are not established.

13. Prevention

There is no evidence-based primary prevention because no modifiable cause is known. No vaccine, prophylactic drug, population screening test, hereditary cascade program, or validated high-risk population exists. Secondary prevention is therefore limited to prompt expert evaluation of suspicious lesions and accurate fusion testing. Tertiary prevention includes surveillance, early management of effusions and organ compromise, pain control, fracture prevention/rehabilitation for bone disease, nutrition support, and avoidance of unnecessary toxic treatment in stable disease.

14. Natural disease in other species

Naturally occurring vascular tumors called hemangioendothelioma have been described in veterinary pathology, but the retrieved evidence does not establish a spontaneous nonhuman disease molecularly homologous to human WWTR1::CAMTA1 EHE. No breed association, VBO term, cross-species transmission, or zoonotic potential is established. Orthologous Wwtr1, Camta1, Yap1, and Tfe3 genes are conserved, but conservation alone does not prove natural EHE equivalence.

15. Model organisms and experimental systems

Conditional knock-in mouse (Mus musculus; NCBI Taxon 10090): Wwtr1-Camta1 expressed from the endogenous locus after Cre activation produces tumors clinically, histologically, immunohistochemically, genetically, and transcriptionally resembling human EHE. No other tumor type was observed, demonstrating fusion sufficiency and tissue specificity. The abstract states that the tumors were “indistinguishable from human EHE clinically, histologically, immunohistochemically, and genetically.” (published March 2021; DOI: https://doi.org/10.1101/gad.348220.120). (seavey2022unravelingthebiology pages 7-8)
Tet-Off mouse: postnatal fusion expression creates hyperplastic pulmonary lesions; doxycycline re-administration causes involution, supporting continued oncogene dependence. Embryonic expression is lethal, limiting developmental modeling. (seavey2022unravelingthebiology pages 7-8)
NIH3T3/TAZ-CAMTA1 cells and xenografts: reproduce anchorage-independent growth and enabled identification of CTGF–integrin αIIbβ3–RAS–MAPK signaling and trametinib sensitivity. Limitation: fibroblast-derived transformed cells are not native human EHE endothelium. (ma2022thetazcamta1fusion pages 1-2)
2024 EHE PDX and paired cell line: derived from aggressive WWTR1::CAMTA1-positive disease and retained original histomorphology plus genomic/transcriptomic profiles. This model enabled sirolimus–doxorubicin comparison and GDF-15 studies. Limitation: one aggressive donor cannot represent the full clinical spectrum. (stacchiotti2024gdf15predictsepithelioid pages 1-2)

No mature zebrafish, Drosophila, organoid, or iPSC model was supported by the retrieved evidence. The mouse and PDX systems are currently the most disease-faithful preclinical resources.

Expert interpretation and evidence gaps

EHE should be treated as a fusion-defined YAP/TAZ-driven sarcoma, not as a low-grade angiosarcoma. Molecular confirmation is central because morphology and clinical tempo overlap with benign vascular tumors, carcinoma, and aggressive angiosarcoma. Surgery can cure localized disease; observation is often appropriate for stable asymptomatic metastatic disease; and mTOR inhibition currently has the strongest recurring disease-specific clinical signal, while MEK and direct YAP/TAZ–TEAD or chromatin-complex inhibition remain rational investigational strategies. (stacchiotti2024gdf15predictsepithelioid pages 1-2, NCT06408441 chunk 1, ma2022thetazcamta1fusion pages 1-2)

The most important unresolved issues are prospective natural-history stratification, validated response criteria for very slow-growing disease, treatment efficacy in serosal/effusion-positive EHE, validation of GDF-15 and CTGF, and molecular distinction of YAP1::TFE3 tumors. The EURACAN registry directly addresses several of these gaps, but its completion is projected for 2033. (NCT06408441 chunk 1)

References

(NCT06408441 chunk 1): Annalisa Trama. The Epithelioid Hemangioendothelioma Registry of the European Reference Network on Rare Adult Solid Cancers (EURACAN). Fondazione IRCCS Istituto Nazionale dei Tumori, Milano. 2023. ClinicalTrials.gov Identifier: NCT06408441
(stacchiotti2024gdf15predictsepithelioid pages 1-2): Silvia Stacchiotti, Silvia Martini, Sandro Pasquali, Anna M. Frezza, Alessia Beretta, Stefano Percio, Mara Lecchi, Monica Tortoreto, Marta Barisella, Paola Collini, Gian Paolo Dagrada, Alessandra Merlini, Paul H. Huang, Andrew Jenks, Robin L. Jones, William D. Tap, Matilde Ingrosso, Carlo Morosi, Silvia Brich, Claudia Giani, Paolo Verderio, Paolo G. Casali, Hugh Leonard, Alessandro Gronchi, Valentina Zuco, and Nadia Zaffaroni. Gdf-15 predicts epithelioid hemangioendothelioma aggressiveness and is downregulated by sirolimus through atf4/atf5 suppression. Sep 2024. URL: https://doi.org/10.1158/1078-0432.ccr-23-3991, doi:10.1158/1078-0432.ccr-23-3991. This article has 13 citations and is from a highest quality peer-reviewed journal.
(seavey2022unravelingthebiology pages 2-4): Caleb Seavey, Ajaybabu Pobbati, and Brian Rubin. Unraveling the biology of epithelioid hemangioendothelioma, a taz–camta1 fusion driven sarcoma. Cancers, 14:2980, Jun 2022. URL: https://doi.org/10.3390/cancers14122980, doi:10.3390/cancers14122980. This article has 22 citations.
(lamar2018epithelioidhemangioendotheliomaas pages 1-3): John M. Lamar, Vijeyaluxmy Motilal Nehru, and Guy Weinberg. Epithelioid hemangioendothelioma as a model of yap/taz-driven cancer: insights from a rare fusion sarcoma. Cancers, 10:229, Jul 2018. URL: https://doi.org/10.3390/cancers10070229, doi:10.3390/cancers10070229. This article has 47 citations.
(lamar2018epithelioidhemangioendotheliomaas pages 5-7): John M. Lamar, Vijeyaluxmy Motilal Nehru, and Guy Weinberg. Epithelioid hemangioendothelioma as a model of yap/taz-driven cancer: insights from a rare fusion sarcoma. Cancers, 10:229, Jul 2018. URL: https://doi.org/10.3390/cancers10070229, doi:10.3390/cancers10070229. This article has 47 citations.
(lamar2018epithelioidhemangioendotheliomaas pages 3-5): John M. Lamar, Vijeyaluxmy Motilal Nehru, and Guy Weinberg. Epithelioid hemangioendothelioma as a model of yap/taz-driven cancer: insights from a rare fusion sarcoma. Cancers, 10:229, Jul 2018. URL: https://doi.org/10.3390/cancers10070229, doi:10.3390/cancers10070229. This article has 47 citations.
(seavey2022unravelingthebiology pages 7-8): Caleb Seavey, Ajaybabu Pobbati, and Brian Rubin. Unraveling the biology of epithelioid hemangioendothelioma, a taz–camta1 fusion driven sarcoma. Cancers, 14:2980, Jun 2022. URL: https://doi.org/10.3390/cancers14122980, doi:10.3390/cancers14122980. This article has 22 citations.
(ong2021genefusionsin pages 13-13): Sheena L. M. Ong, Karoly Szuhai, and Judith V.M.G. Bovée. Gene fusions in vascular tumors and their underlying molecular mechanisms. Expert Review of Molecular Diagnostics, 21:897-909, Aug 2021. URL: https://doi.org/10.1080/14737159.2021.1950533, doi:10.1080/14737159.2021.1950533. This article has 14 citations and is from a peer-reviewed journal.
(stacchiotti2024gdf15predictsepithelioid pages 9-9): Silvia Stacchiotti, Silvia Martini, Sandro Pasquali, Anna M. Frezza, Alessia Beretta, Stefano Percio, Mara Lecchi, Monica Tortoreto, Marta Barisella, Paola Collini, Gian Paolo Dagrada, Alessandra Merlini, Paul H. Huang, Andrew Jenks, Robin L. Jones, William D. Tap, Matilde Ingrosso, Carlo Morosi, Silvia Brich, Claudia Giani, Paolo Verderio, Paolo G. Casali, Hugh Leonard, Alessandro Gronchi, Valentina Zuco, and Nadia Zaffaroni. Gdf-15 predicts epithelioid hemangioendothelioma aggressiveness and is downregulated by sirolimus through atf4/atf5 suppression. Sep 2024. URL: https://doi.org/10.1158/1078-0432.ccr-23-3991, doi:10.1158/1078-0432.ccr-23-3991. This article has 13 citations and is from a highest quality peer-reviewed journal.
(lamar2018epithelioidhemangioendotheliomaas pages 7-8): John M. Lamar, Vijeyaluxmy Motilal Nehru, and Guy Weinberg. Epithelioid hemangioendothelioma as a model of yap/taz-driven cancer: insights from a rare fusion sarcoma. Cancers, 10:229, Jul 2018. URL: https://doi.org/10.3390/cancers10070229, doi:10.3390/cancers10070229. This article has 47 citations.
(NCT03148275 chunk 3): Trametinib in Treating Patients With Epithelioid Hemangioendothelioma That is Metastatic, Locally Advanced, or Cannot Be Removed by Surgery. National Cancer Institute (NCI). 2017. ClinicalTrials.gov Identifier: NCT03148275
(stacchiotti2024gdf15predictsepithelioid pages 11-12): Silvia Stacchiotti, Silvia Martini, Sandro Pasquali, Anna M. Frezza, Alessia Beretta, Stefano Percio, Mara Lecchi, Monica Tortoreto, Marta Barisella, Paola Collini, Gian Paolo Dagrada, Alessandra Merlini, Paul H. Huang, Andrew Jenks, Robin L. Jones, William D. Tap, Matilde Ingrosso, Carlo Morosi, Silvia Brich, Claudia Giani, Paolo Verderio, Paolo G. Casali, Hugh Leonard, Alessandro Gronchi, Valentina Zuco, and Nadia Zaffaroni. Gdf-15 predicts epithelioid hemangioendothelioma aggressiveness and is downregulated by sirolimus through atf4/atf5 suppression. Sep 2024. URL: https://doi.org/10.1158/1078-0432.ccr-23-3991, doi:10.1158/1078-0432.ccr-23-3991. This article has 13 citations and is from a highest quality peer-reviewed journal.
(OpenTargets Search: epithelioid hemangioendothelioma): Open Targets Query (epithelioid hemangioendothelioma, 10 results). Buniello, A. et al. (2025). Open Targets Platform: facilitating therapeutic hypotheses building in drug discovery. Nucleic Acids Research.
(ma2022thetazcamta1fusion pages 1-2): Shuang Ma, Ryan Kanai, Ajaybabu V. Pobbati, Shuo Li, Kepeng Che, Caleb N. Seavey, Andrea Hallett, Ashley Burtscher, John M. Lamar, and Brian P. Rubin. The taz-camta1 fusion protein promotes tumorigenesis via connective tissue growth factor and ras–mapk signaling in epithelioid hemangioendothelioma. Clinical Cancer Research, 28:3116-3126, Apr 2022. URL: https://doi.org/10.1158/1078-0432.ccr-22-0421, doi:10.1158/1078-0432.ccr-22-0421. This article has 37 citations and is from a highest quality peer-reviewed journal.
(ong2021genefusionsin pages 8-10): Sheena L. M. Ong, Karoly Szuhai, and Judith V.M.G. Bovée. Gene fusions in vascular tumors and their underlying molecular mechanisms. Expert Review of Molecular Diagnostics, 21:897-909, Aug 2021. URL: https://doi.org/10.1080/14737159.2021.1950533, doi:10.1080/14737159.2021.1950533. This article has 14 citations and is from a peer-reviewed journal.
(NCT03148275 chunk 1): Trametinib in Treating Patients With Epithelioid Hemangioendothelioma That is Metastatic, Locally Advanced, or Cannot Be Removed by Surgery. National Cancer Institute (NCI). 2017. ClinicalTrials.gov Identifier: NCT03148275
(wiegand2021malignantvasculartumors pages 7-9): Susanne Wiegand, Andreas Dietz, and Gunnar Wichmann. Malignant vascular tumors of the head and neck—which type of therapy works best? Cancers, 13:6201, Dec 2021. URL: https://doi.org/10.3390/cancers13246201, doi:10.3390/cancers13246201. This article has 16 citations.
(NCT03148275 chunk 2): Trametinib in Treating Patients With Epithelioid Hemangioendothelioma That is Metastatic, Locally Advanced, or Cannot Be Removed by Surgery. National Cancer Institute (NCI). 2017. ClinicalTrials.gov Identifier: NCT03148275
(NCT03331250 chunk 1): Gregory Cote. Eribulin in Angiosarcoma and Epithelioid Hemangioendothelioma (EHE). Massachusetts General Hospital. 2018. ClinicalTrials.gov Identifier: NCT03331250

Artifacts

Edison artifact artifact-00

Ask OpenScientist

OpenScientist Report

Classifications

Subtypes

Discussions and Knowledge Gaps

Pathophysiology

Histopathology

Pathograph

Phenotypes

Genetic Associations

Medical Actions

Biochemical Markers

Progression

Prevalence

Differential Diagnoses

Clinical Trials

Experimental Models

Source YAML

References & Deep Research

Deep Research

Epithelioid Hemangioendothelioma: Comprehensive Disease-Characteristics Report

1. Disease information

Definition

Identifiers and terminology

2. Etiology, risk, and protective factors

Causal lesion

Risk and protective factors

3. Phenotypes and quality-of-life effects

4. Genetic and molecular information

Principal alterations

Epigenetic/chromatin biology

5. Environmental and infectious information

6. Mechanism and pathophysiology

Causal chain

mTOR and GDF-15

Molecular-profile and ontology suggestions

7. Anatomy

8. Temporal development and natural history

9. Inheritance and population epidemiology

10. Diagnostics

Recommended workflow

Differential diagnosis

11. Outcome and prognosis

12. Treatment and real-world implementation

Strategy

Systemic therapies

13. Prevention

14. Natural disease in other species

15. Model organisms and experimental systems

Expert interpretation and evidence gaps

Artifacts